MultiFinBen is TheFinAI's multilingual, multimodal financial benchmark (FINOS-backed, Linux Foundation), spanning financial text, vision, and audio tasks. AgMoDB records the official modality-balanced aggregate published by the current leaderboard.
This definition has not completed editorial and provenance review. Verify its metric, version, source, and evaluation configuration upstream before using it for a model-selection decision.
Industry-backed and modality-diverse; covers multilingual financial tasks and includes open-source and proprietary models; the balanced aggregate prevents a modality with more tasks from dominating the headline score.
The current leaderboard publishes its canonical aggregate in frontend source while its legacy formatted API is unreliable. AgMoDB publishes scores only for models already represented in its decision catalog and logs excluded source rows. Per-modality averages can hide uneven task performance, and this finance-specific benchmark is not a proxy for general capability.
Higher is better. Modality-balanced score on a 0–100 scale, combining the leaderboard's text, vision, and audio averages.
The current leaderboard publishes its canonical aggregate in frontend source while its legacy formatted API is unreliable.
| # | Model | Score |
|---|---|---|
| 1 | 48.3 | |
| 2 | 46.0 | |
| 3 | 43.4 | |
| 4 | 42.5 | |
| 5 | 41.4 | |
| 6 | 31.2 | |
| 7 | 28.3 | |
| 8 | 23.2 | |
| 9 | 22.7 | |
| 10 | 22.7 | |
| 11 | 20.9 | |
| 12 | 19.8 | |
| 13 | 18.9 | |
| 14 | 17.4 | |
| 15 | 15.7 | |
| 16 | 14.7 | |
| 17 | 14.3 | |
| 18 | 14.1 | |
| 19 | 14.1 | |
| 20 | 12.7 | |
| 21 | 10.2 | |
| 22 | 8.9 | |
| 23 | 8.9 | |
| 24 | 8.1 | |
| 25 | 7.8 | |
| 26 | 4.9 |