AgMoDB
ModelsAgentsOrchestratorsEvalsCompositesVisualizeIndustry
AgMoDB by @mistakeknot
All Models

Gemini 3.8 Flash (High)

September 2, 2026Last synced Nov 22, 2026, 8:17 AM1049K context
Effort
  • Low
  • Medium
  • High

Blended Price

$1.50/M

Input Price

$0.75/M

Output Price

$3.75/M

Speed

114 tok/s

TTFT

19.16s

Benchmark Scores

Intelligence Index
30.8
Coding Index
—
Math Index
—
MMLU Pro
— / 100
GPQA Diamond
95.3 / 100
HLE
47.8 / 100
LiveCodeBench
— / 100
SciCode
56.6 / 100
MATH-500
— / 100
AIME
— / 30
AIME 2025
— / 30
IFBench
— / 100
LCR
81.3 / 100
Terminal-Bench Hard
— / 100
τ²-Bench
— / 100

Performance Radar

Gemini 3.8 Flash (High)
Field Average

Hidden, no data: AgMoBench Agentic, AgMoBench Robustness

External Benchmarks

AA-Omniscience Accuracyaa_omniscience
50.5 / 100
AA-Omniscience Hallucination Rateaa_omniscience
55.1 / 100
AA-Omniscience Indexaa_omniscience
23.3 / 100
ARC-AGI-1arc_agi
98.5 / 100
ARC-AGI-1 Cost per Taskarc_agi
0.2
ARC-AGI-2arc_agi
89.2 / 100
ARC-AGI-2 Cost per Taskarc_agi
0.4
Arena ELO: Codingchatbot_arena
1584.0
arena_elo_documentchatbot_arena
1482.0
Chatbot Arena ELOchatbot_arena
1500.0
arena_elo_visionchatbot_arena
1316.0
Ale Benchepoch_ai
1269.7
Arc Agi 2epoch_ai
0.9
Chess Puzzlesepoch_ai
0.6
Critptepoch_ai
0.2
Cursorbenchepoch_ai
0.4
DeepSWE (Epoch Import, Version Unverified)epoch_ai
0.7 / 100
Eci Scoresepoch_ai
161.3
Frontiermath Erdosepoch_ai
0.0
Frontiermath Tier 4 V2epoch_ai
0.2
Frontiermath Tiers 1 3 V2epoch_ai
0.7
Frontiersweepoch_ai
0.2
Furniture Assemblyepoch_ai
0.3
Gdp Pdfepoch_ai
0.4
Mirrorcodeepoch_ai
0.4
Mystery Game Puzzlesepoch_ai
0.5
OTIS Mock AIME 2024–2025epoch_ai
1.0
Processed Data For Eciepoch_ai
0.4
Scicodeepoch_ai
0.6
SimpleQA Verified (Epoch)epoch_ai
0.7
Surface Evolver Benchepoch_ai
0.8
WebDev Arenaepoch_ai
1567.9
Weirdml V3epoch_ai
0.1
LegalBenchlegalbench
87.0 / 100
LiveBench Codinglivebench
73.3 / 100
LiveBench Data Analysislivebench
54.0 / 100
LiveBench Instruction Followinglivebench
81.4 / 100
LiveBench Languagelivebench
87.8 / 100
LiveBench Mathlivebench
91.6 / 100
LiveBench Reasoninglivebench
89.3 / 100
OpenRouter Token Volumeopenrouter_rankings
1750000000000.0
OpenRouter Weekly Trendopenrouter_rankings
14.0
WeirdMLweirdml
84.8 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"ec09fbed9f211f59ec54cca2227c3309\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"5814b0938ba4a54ba31da46fe17d4380\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-20

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"ec09fbed9f211f59ec54cca2227c3309\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"5814b0938ba4a54ba31da46fe17d4380\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-20

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"d8dfdc9958a2e1502b54603b03fbc557\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"aaa10807f3dcaae3ce6c15be7539e733\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-20

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"d8dfdc9958a2e1502b54603b03fbc557\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"aaa10807f3dcaae3ce6c15be7539e733\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-20

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"f4d8307b32b2fefe010860bedd0c5dd8\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ef15ebd7644a875688eda094d716da9a\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-18

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"f4d8307b32b2fefe010860bedd0c5dd8\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ef15ebd7644a875688eda094d716da9a\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-18

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"f4d8307b32b2fefe010860bedd0c5dd8\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ef15ebd7644a875688eda094d716da9a\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-17

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"f4d8307b32b2fefe010860bedd0c5dd8\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ef15ebd7644a875688eda094d716da9a\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-17

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"69aae253e532553a920bf4deb50ca495\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"5206202f381a21893de82dcc2db93ae9\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-16

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"69aae253e532553a920bf4deb50ca495\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"5206202f381a21893de82dcc2db93ae9\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-16

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"513aaaa353972ca0baa9c95cb95e04dd\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"169653a17d103c8dbca682167ffccaa6\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-10

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"513aaaa353972ca0baa9c95cb95e04dd\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"169653a17d103c8dbca682167ffccaa6\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-10

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"3d7dcca4fb88130b9e7ec186fd9ae23f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"1e46e6ab8c83fb1a7b19f10be398adf2\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-09

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"3d7dcca4fb88130b9e7ec186fd9ae23f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"1e46e6ab8c83fb1a7b19f10be398adf2\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-09

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8ea4d2292a0a8c759aa7e80a2041122d\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"94fd3d68aacca744ff3c39f63a187685\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-06

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8ea4d2292a0a8c759aa7e80a2041122d\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"94fd3d68aacca744ff3c39f63a187685\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-06

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"c4baed683b2f33c40537f2338682475f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"48d1fd5bf96ae9583213f6699f12f2cc\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-04

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"c4baed683b2f33c40537f2338682475f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"48d1fd5bf96ae9583213f6699f12f2cc\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-04

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"c4baed683b2f33c40537f2338682475f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"48d1fd5bf96ae9583213f6699f12f2cc\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-02

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"c4baed683b2f33c40537f2338682475f\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"48d1fd5bf96ae9583213f6699f12f2cc\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-11-02

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8fc20a5217998d26c40ad2306d648194\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"095b2fa20aac96121092dab07e18cc8e\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-28

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8fc20a5217998d26c40ad2306d648194\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"095b2fa20aac96121092dab07e18cc8e\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-28

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"400b32b66e8c1b37fc50ff586b033048\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"9314d4bea0d9d001c69a11614dc8d93c\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-21

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"400b32b66e8c1b37fc50ff586b033048\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"9314d4bea0d9d001c69a11614dc8d93c\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-21

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"2ab08f641c16bcf3c7a59c38b35fa2e3\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"b4ed41aee6b00420557f098bd05f4165\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-19

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"2ab08f641c16bcf3c7a59c38b35fa2e3\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"b4ed41aee6b00420557f098bd05f4165\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-19

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8cf5cb76df0b13c1684b346a2012df83\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"c6e9a78d12dc070d7dab18a1f667a410\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-15

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8cf5cb76df0b13c1684b346a2012df83\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"c6e9a78d12dc070d7dab18a1f667a410\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-15

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"57cc4a7689e563a0ffb7d275eb3a889b\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"19d71f8c9a2056c0be4bb4fd4a75a325\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-14

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"57cc4a7689e563a0ffb7d275eb3a889b\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"19d71f8c9a2056c0be4bb4fd4a75a325\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-14

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"57cc4a7689e563a0ffb7d275eb3a889b\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"19d71f8c9a2056c0be4bb4fd4a75a325\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-14

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"57cc4a7689e563a0ffb7d275eb3a889b\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"19d71f8c9a2056c0be4bb4fd4a75a325\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-14

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"e7f673a6db195dac7f2956e3b1651145\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"4d3b46c6ffe9404ca1c0ebcfb5938883\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-13

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"e7f673a6db195dac7f2956e3b1651145\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"4d3b46c6ffe9404ca1c0ebcfb5938883\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-13

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b57a26ca4df83acaa40eadb6632c7c61\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ee35c9dbbb033dc7de85467eedb85d54\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-06

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b57a26ca4df83acaa40eadb6632c7c61\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"ee35c9dbbb033dc7de85467eedb85d54\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-06

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b52a6e82589fb2c719168d69aa0bb015\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"f187ceaff4d7f6156e77c1f1254aafc8\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-02

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b52a6e82589fb2c719168d69aa0bb015\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"f187ceaff4d7f6156e77c1f1254aafc8\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-02

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b52a6e82589fb2c719168d69aa0bb015\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"f187ceaff4d7f6156e77c1f1254aafc8\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-01

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"b52a6e82589fb2c719168d69aa0bb015\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"f187ceaff4d7f6156e77c1f1254aafc8\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-10-01

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8bfe2e142f29acec39df406a6631f81a\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"92bfaa47028eae49163960f4c5a298a0\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-30

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"8bfe2e142f29acec39df406a6631f81a\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"92bfaa47028eae49163960f4c5a298a0\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-30

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"16793624264c3bdded076b4e26af5a90\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"e7fb37556a7353e7da435c4c7b78ca50\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-29

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"16793624264c3bdded076b4e26af5a90\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"e7fb37556a7353e7da435c4c7b78ca50\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-29

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"284ac3aff89a7d8a3aca179ee1e5dd61\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"6e0849a05a2c572b1050e5900a372e2b\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-28

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"284ac3aff89a7d8a3aca179ee1e5dd61\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"6e0849a05a2c572b1050e5900a372e2b\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-28

10.4 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash - Provider Adapter (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high-provider-adapter

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash-provider-adapter",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the provider adapter harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"397d7634d348db2ae8df5180db7b475d\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"882404aedf04dbdde2b6c0d731212a4f\"",
  "upstreamModelId": "google-gemini-3-8-flash-high-provider-adapter",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4521.7983525,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-24

35.0 / 100
ARC-AGI-3 Public Demoarc_agi

v3 · public-demo · Gemini 3.8 Flash (High) · effort high · 1 attempt · 5x-human-baseline-median-per-level · arc-agi-3:public-demo · record v3_Semi_Private:google-gemini-3-8-flash-high

View run source

Source metadata
{
  "provider": "Google",
  "modelType": "CoT",
  "modelsUrl": "https://arcprize.org/media/data/models.json",
  "modelGroup": "google-gemini-3-8-flash",
  "modelNotes": "Evaluation conducted with 'high' reasoning effort using the standard harness. Partial dataset coverage: scores use full catalog denominators; costs include recorded usage only, excluding unrecorded attempts.",
  "modelsEtag": "W/\"397d7634d348db2ae8df5180db7b475d\"",
  "scoreMetric": "relative-human-action-efficiency",
  "evaluationsUrl": "https://arcprize.org/media/data/evaluations.json",
  "evaluationsEtag": "W/\"882404aedf04dbdde2b6c0d731212a4f\"",
  "upstreamModelId": "google-gemini-3-8-flash-high",
  "environmentCount": 25,
  "modelReleaseDate": "2026-09-02T00:00:00.000Z",
  "upstreamDatasetId": "v3_Semi_Private",
  "actionBudgetPolicy": "5x-human-baseline-median-per-level",
  "technicalReportUrl": "https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf",
  "catalogMatchAliases": [],
  "agentHarnessPublished": false,
  "verificationAuthority": "ARC Prize Foundation",
  "upstreamRunIdPublished": false,
  "actualEvaluationCostUsd": 4398.2756895,
  "providerModelIdPublished": false
}

Evidence observed 2026-09-24

10.4 / 100

How Gemini 3.8 Flash (High) Compares

Axes

Filters

$0.00 – $67.5
≥ 1.4

Providers

ai2ai21-labsalibabaanthropicarceeawsazurebaidu
openai
zai
meta
nvidia
xai
upstage
deepseek
alibaba
mistral
nous-research
aws
google
anthropic
xiaomi
stepfun
kimi
reflection
minimax
azure
inclusionai
reka-ai
cohere
bytedance_seed
inception
tencent
thinking-machines
Bubble size = Context Window
8191.00
2000000.00
Compare with other models