Forecastbench

Forecastbench benchmark scores from Epoch AI's benchmark data collection.

aggregateepoch_ai

Strengths

Sourced from Epoch AI's systematic benchmark data collection covering frontier models.

Caveats

Auto-generated definition — editorial description not yet written.

How to interpret scores

Contamination risk: moderateFreshness: periodic

Higher is better.

Caveat to keep in mind

Auto-generated definition — editorial description not yet written.

Leaderboard

Last synced: Jul 10, 2026, 8:59 PM

#	Model	Provider	Score	Price/M
1	Gemini 2.5 Flash (Non-reasoning)	Google	1.0	$0.85
2	Qwen2.5 72B Instruct	Alibaba	1.0	$0.19
3	Google: Gemini 3 Flash Preview	Google	1.0	$1.13
4	Claude 3 Opus	Anthropic	1.0	$30.00
5	QwQ 32B-Preview	Alibaba	1.0	$0.00
6	DeepSeek V3.1 (Non-reasoning)	DeepSeek	1.0	$0.84
7	GPT-4	OpenAI	1.0	$37.50
8	GPT-4o (May '24)	OpenAI	1.0	$7.50
9	Llama 4 Scout	Meta	1.0	$0.29
10	Claude 2.1	Anthropic	1.0	$0.00
11	Claude 3 Haiku	Anthropic	1.0	$0.50
12	GPT-3.5 Turbo	OpenAI	1.0	$0.75
13	Llama 4 Maverick	Meta	1.0	$0.47
14	Mistral Large 2 (Jul '24)	Mistral	1.0	$3.00
15	Meta: Llama 3 70B Instruct	Meta	1.0	$0.57
16	Mistral Large 2411	Mistral	1.0	$3.00
17	Mixtral 8x7B Instruct	Mistral	1.0	$0.51
18	Mixtral 8x22B Instruct	Mistral	1.0	$0.00
19	Gemini 3.1 Flash-Lite	Google	1.0	$0.56
20	Llama 3.1 Instruct 405B	Meta	1.0	$3.69
21	Qwen: Qwen3 235B A22B	Alibaba	1.0	$0.39
22	o3-mini	OpenAI	1.0	$1.93
23	GPT-4 Turbo	OpenAI	1.0	$15.00
24	Claude 3.5 Sonnet (Oct '24)	Anthropic	1.0	$6.00
25	GLM-4.5-Air	Z AI	1.0	$0.37
26	DeepSeek V3 (Dec '24)	DeepSeek	1.0	$0.52
27	Gemini 3.1 Pro Preview	Google	1.0	$4.50
28	Meta: Llama 3.3 70B Instruct	Meta	1.0	$0.15
29	Kimi K2	Kimi	1.0	$1.04
30	Meta: Llama 3 8B Instruct	Meta	1.0	$0.03
31	Grok 4 Fast (Non-reasoning)	SpaceXAI	1.0	$0.28
32	Anthropic: Claude Sonnet 4	Anthropic	1.0	$6.00
33	DeepSeek R1 (Jan '25)	DeepSeek	1.0	$2.43
34	Gemini 2.5 Pro	Google	1.0	$3.44
35	Claude Opus 4.5 (Non-reasoning)	Anthropic	1.0	$10.00
36	Grok 4.1 Fast (Reasoning)	SpaceXAI	1.0	$0.00
37	Grok 4	SpaceXAI	1.0	$11.00
38	Gemini 3 Pro Preview (low)	Google	1.0	$4.50
39	GLM-5 (Reasoning)	Z AI	1.0	$1.55
40	Claude Opus 4.7	Anthropic	0.9	$11.67
41	Anthropic: Claude Haiku 4.5	Anthropic	0.9	$2.00
42	GPT-4.1	OpenAI	0.9	$3.50
43	Claude 3.7 Sonnet (Non-reasoning)	Anthropic	0.9	$6.00
44	Grok 4.20 0309 v2 (Reasoning)	SpaceXAI	0.9	$3.00
45	GPT-4.5 (Preview)	OpenAI	0.9	$0.00
46	Anthropic: Claude Sonnet 4.5	Anthropic	0.9	$6.00
47	Anthropic: Claude Opus 4.1	Anthropic	0.9	$30.00
48	o3	OpenAI	0.8	$3.50