| # ↕ | Model ↕ | FB Score ↓ | PAVE ↕ | FinanceBench ↕ | FinQA ↕ | FinAuditing ↕ | FinBen ↕ | Benchmarks ↕ | Released ↕ | License | Context | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 🥇 | Anthropic | 73.8 | 73.0 | 88.0 | 84.0 | 75.0 | 81.0 | 10 / 42 | 2026-03 | Prop. | 200K | |
| 🥈 | OpenAI | 72.2 | 71.0 | 91.0 | 82.0 | 72.0 | 79.0 | 12 / 42 | 2026-02 | Prop. | 128K | |
| 3 | OpenAI | 71.1* | — | — | — | — | — | 1 benchmark 1 / 42 | 2026-09-16 | Prop. | 1M | |
| 🥉 | DeepSeek | 70.1 | 65.0 | 85.0 | 87.0 | 73.0 | 78.0 | 9 / 42 | 2026-01 | Open | 128K | |
| 5 | Google | 70.0 | 69.0 | 86.0 | 80.0 | 70.0 | 85.0 | 9 / 42 | 2026-01 | Prop. | 1M | |
| 6 | OpenAI | 70.0* | — | — | — | — | — | 1 benchmark 1 / 42 | 2026-09-20 | Prop. | 1M | |
| 7 | Anthropic | 69.8 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-06-09 | Prop. | 1M | |
| 8 | Moonshot AI | 67.4 | 66.0 | 82.0 | 76.0 | — | 75.0 | 9 / 42 | 2026-02 | Open | 128K | |
| 9 | Anthropic | 66.7 | — | — | — | — | — | 30 benchmarks 30 / 42 | 2026-07-22 | Prop. | 1M | |
| 10 | Anthropic | 66.2 | — | — | — | — | — | 5 benchmarks 5 / 42 | 2026-09-01 | Prop. | 1M | |
| 11 | Zhipu AI | 65.9 | 63.0 | 79.0 | 74.0 | — | 73.0 | 7 / 42 | 2026-01 | Open | 128K | |
| 12 | xAI | 65.4 | — | — | — | — | — | 21 benchmarks 21 / 42 | 2026-08-12 | Prop. | 500K | |
| 13 | OpenAI | 65.1 | — | — | — | — | — | 29 benchmarks 29 / 42 | 2026-07-09 | Prop. | 1M | |
| 14 | xAI | 65.0 | 68.0 | 84.0 | 78.0 | — | 77.0 | 9 / 42 | 2026-02 | Prop. | 131K | |
| 15 | Google | 64.8 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-02 | Prop. | 1M | |
| 16 | OpenAI | 64.6 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-07-09 | Prop. | 1M | |
| 17 | Anthropic | 64.1 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-06-30 | Prop. | 1M | |
| 18 | DeepSeek | 63.7 | — | — | — | — | — | 16 benchmarks 16 / 42 | 2026-08-13 | Open | 1M | |
| 19 | Meta | 63.7 | — | — | — | — | — | 23 benchmarks 23 / 42 | 2026-08-05 | Prop. | 1M | |
| 20 | Moonshot AI | 63.4 | — | — | — | — | — | 27 benchmarks 27 / 42 | 2026-07-16 | Open | 1M | |
| 21 | Zhipu AI | 63.4 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-08 | Open | 1M | |
| 22 | DeepSeek | 63.2 | — | — | — | — | — | 19 benchmarks 19 / 42 | 2026-07-31 | Open | 1M | |
| 23 | OpenAI | 63.1 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-07-09 | Prop. | 1M | |
| 24 | Google | 63.1 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-10-13 | Prop. | 1M | |
| 25 | Meta | 62.8 | — | — | — | — | — | 23 benchmarks 23 / 42 | 2026-07-09 | Prop. | 1M | |
| 26 | Alibaba | 62.2 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-08-03 | Open | 1M | |
| 27 | Google | 61.6 | — | — | — | — | — | 22 benchmarks 22 / 42 | 2026-07-21 | Prop. | 1M | |
| 28 | Google | 61.2 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-08 | Prop. | 1M | |
| 29 | Anthropic | 59.4 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-22 | Prop. | 1M | |
| 30 | OpenAI | 59.3 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-03 | Prop. | 1.1M | |
| 31 | MiniMax | 58.7 | 62.0 | 77.0 | 72.0 | — | 71.0 | 8 / 42 | 2026-01 | Open | 1M | |
| 32 | Anthropic | 58.4 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-18 | Prop. | 1M | |
| 33 | Meta | 58.0 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-02 | Prop. | 1M | |
| 34 | Zhipu AI | 57.5 | — | — | — | — | — | 20 benchmarks 20 / 42 | 2026-06-13 | Open | 1M | |
| 35 | MiniMax | 57.2 | — | — | — | — | — | 5 benchmarks 5 / 42 | 2026-06 | Open | 1M | |
| 36 | xAI | 56.8 | — | — | — | — | — | 25 benchmarks 25 / 42 | 2026-07-16 | Prop. | 500K | |
| 37 | Zhipu AI | 55.3 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-08 | Open | 1M | |
| 38 | Meta | 55.1 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-02 | Prop. | 1M | |
| 39 | DeepSeek | 54.8 | — | — | — | — | — | 3 benchmarks 3 / 42 | 2026-09-10 | Open | 1M | |
| 40 | Google | 53.5 | — | — | — | — | — | 23 benchmarks 23 / 42 | 2026-07-21 | Prop. | 1M | |
| 41 | Moonshot AI | 53.5 | — | — | — | — | — | 7 benchmarks 7 / 42 | 2026-06-12 | Open | 262K | |
| 42 | Thinking Machines | 52.3 | — | — | — | — | — | 26 benchmarks 26 / 42 | 2026-07-15 | Open | 256K | |
| 43 | Thinking Machines | 51.5 | — | — | — | — | — | 24 benchmarks 24 / 42 | 2026-07-30 | Open | 262K | |
| 44 | NVIDIA | 46.3 | — | — | — | — | — | 16 benchmarks 16 / 42 | 2026-06-04 | Open | 1M | |
| 45 | Ant Group | 45.0 | — | — | — | — | — | 18 benchmarks 18 / 42 | 2026-07-23 | Open | 262K | |
| 46 | Alibaba | 39.5 | — | — | — | — | — | 12 benchmarks 12 / 42 | 2026-06-01 | Prop. | 1M | |
| 47 | Zhipu AI | —* | — | — | — | — | — | 0 benchmarks 0 / 42 | 2026-08 | Open | 128K | |
| 48 | Google | —* | — | — | — | — | — | 0 benchmarks 0 / 42 | 2026-09-24 | Prop. | 1M | |
| 49 | Anthropic | —* | — | — | — | — | — | 0 benchmarks 0 / 42 | 2026-10-06 | Prop. | 1M | |
| 50 | Anthropic | —* | — | — | — | — | — | 0 benchmarks 0 / 42 | 2026-09-15 | Prop. | 1M | |
| 51 | OpenAI | —* | — | — | — | — | — | 0 benchmarks 0 / 42 | 2026-09-25 | Prop. | 128K |
Models by organisation
Anthropic9
OpenAI8
Google7
DeepSeek4
xAI3
Moonshot AI3
Zhipu AI5
MiniMax2
Thinking Machines2
Ant Group1
Meta4
NVIDIA1
Alibaba2
Top LLMs - FB Score
1Claude Opus 4.7
Anthropic · 10/8 benchmarks
GPT-5.5
OpenAI · 12/8 benchmarks
GPT-6 Sol
OpenAI · 1/8 benchmarks
DeepSeek V4 Pro
DeepSeek · 9/8 benchmarks
Gemini 3.1 Pro
Google · 9/8 benchmarks
Open-weight models
19 models in this evaluation are publicly available for download and self-hosting.
DeepSeek V4 Pro70.1%
Kimi K2.667.4%
GLM-5.165.9%
DeepSeek V4 Pro (0813)63.7%
Kimi K363.4%
GLM 5.3 Flash63.4%
DeepSeek V4 Flash63.2%
Qwen 3.8 Max62.2%
MiniMax M2.758.7%
GLM 5.257.5%
MiniMax M357.2%
GLM 5.355.3%
DeepSeek V4.1 Flash54.8%
Kimi K2.7 Code53.5%
Inkling52.3%
Inkling Small51.5%
Nemotron 3 Ultra46.3%
Ling 3.0 Flash45.0%
GLM-5.50.0%
Missing a model?
Submit a score for any model against our tracked benchmarks, or request a certified evaluation via the open harness.
Submit a score →