Open FinLLM Leaderboard: leaderboard
Open financial-language-model leaderboard from FINOS, covering broad financial NLP and reasoning task categories.
Metric: Average normalized score (self-reported). Source: benchmarklist.com. Status: saturated. 22 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | plutus-8B-instruct | 55.9 |
| 2 | GPT-4 | 48.34 |
| 3 | Qwen 2.5 32B Instruct | 42.49 |
| 4 | Qwen 2.5 72B Instruct | 41.36 |
| 5 | GPT-4o | 37.51 |
| 6 | Llama 3 70B Instruct | 31.23 |
| 7 | DeepSeek V3 Chat | 29.49 |
| 8 | GPT-4o Mini | 28.32 |
| 9 | GPT-3.5 Turbo (0125) | 23.16 |
| 10 | Llama 3.1 8B Instruct | 22.72 |
| 11 | Gemma 2 27B (IT) | 19.81 |
| 12 | Llama 3 8B Instruct | 18.93 |
| 13 | Qwen 2.5 7B Instruct | 17.37 |
| 14 | Mistral 7B Instruct (v0.3) | 14.68 |
| 15 | Gemma 2 9B (IT) | 14.32 |
Interactive version: theaggregate.ai/benchmark?slug=open-finllm-leaderboard · How It Works · Data refreshed daily, snapshot 2026-09-05.