Best AI for Finance knowledge (November 2026)
Data refreshed 1 Nov 2026
ChatGPT-o3 leads BizFinBench with 73.9 (average score across 9 tasks). GPT-4o leads Open FinLLM Leaderboard with 46.0 (balanced score (mean of text/vision/audio averages)). Qwen 3.8 Max 0902 leads Tax Agent Bench with 49.7 (accuracy % (all-pass x citation factor)). o4-mini leads Advanced Financial Reasoning at Scale (CFA Level III evaluation) with 79.1 (overall % (mean of MCQ and essay)). Gemini 3.0 Pro leads Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation) with 97.6 (Level I accuracy %).
- Benchmarks in this use case
- 7
- With dated scores
- 5
- Models with scores
- 42
- Latest source date
- 29 Oct 2026
Finance knowledge picks
The leader on each benchmark, never combinedBest on BizFinBench
ChatGPT-o3OpenAI
73.9 average score across 9 tasks
Reported by the authors, 26 May 2025
Best on Open FinLLM Leaderboard
GPT-4oOpenAI
46.0 balanced score (mean of text/vision/audio averages)
Official board, 16 Apr 2026
Best on Tax Agent Bench
Qwen 3.8 Max 0902Qwen
49.7 accuracy % (all-pass x citation factor)
Official board, 29 Oct 2026
Best on Advanced Financial Reasoning at Scale (CFA Level III evaluation)
o4-miniOpenAI
79.1 overall % (mean of MCQ and essay)
Reported by the authors, 22 Sept 2025
Best on Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation)
Gemini 3.0 ProGoogle
97.6 Level I accuracy %
Reported by the authors, 9 Dec 2025
Finance knowledge leaderboard
BizFinBench v1 (paper 2025-05-26), marked outdated by the maker; BizFinBench.v2 released 2026-01-09 (29,578 questions, ICML 2026)
average score across 9 tasks, higher is better. Top 5 of 10 models.
Updated 26 May 2025
- 1ChatGPT-o373.9
- 2DeepSeek-R1 (671B)73.0
- 3GPT-4o71.8
- 4DeepSeek-V3 (671B)71.6
- 5ChatGPT-o4-mini71.3
| 1 | ChatGPT-o3 | OpenAI | 73.9 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
|---|---|---|---|---|---|
| 2 | DeepSeek-R1 (671B) | DeepSeek | 73.0 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 3 | GPT-4o | OpenAI | 71.8 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 4 | DeepSeek-V3 (671B) | DeepSeek | 71.6 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 5 | ChatGPT-o4-mini | OpenAI | 71.3 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 6 | Gemini-2.0-Flash | 69.8 | Default | Paper Table 3 Reported by the authors, 26 May 2025 | |
| 7 | Qwen2.5-VL-32B | Qwen | 68.6 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 8 | Qwen3-32B | Qwen | 68.3 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 9 | Qwen2.5-72B-Instruct | Qwen | 67.7 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
| 10 | Qwen2.5-VL-72B | Qwen | 67.5 | Default | Paper Table 3 Reported by the authors, 26 May 2025 |
Open FinLLM Leaderboard Board data updated 2026-04-16 (MultiFinBen arXiv 2506.14028v3, Table 3)
balanced score (mean of text/vision/audio averages), higher is better. Top 5 of 10 models.
Updated 16 Apr 2026
- 1GPT-4o46.0
- 2Qwen2.5-Omni-7B33.5
- 3Llama-4-Scout-17B-16E20.9
- 4Whisper-V317.2
- 5Qwen2-Audio-7B-Instruct16.7
| 1 | GPT-4o | OpenAI | 46.0 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
|---|---|---|---|---|---|
| 2 | Qwen2.5-Omni-7B | Qwen | 33.5 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 3 | Llama-4-Scout-17B-16E | Meta | 20.9 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 4 | Whisper-V3 | OpenAI | 17.2 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 5 | Qwen2-Audio-7B-Instruct | Qwen | 16.7 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 6 | Qwen2-Audio-7B | Qwen | 16.0 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 7 | Gemma-3-27B | 15.7 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 | |
| 8 | Llama-3.1-70B | Meta | 14.1 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
| 9 | Gemma-3-4B | 12.7 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 | |
| 10 | Qwen2.5-32B | Qwen | 12.7 | MultiFinBen Table 3 data | Open FinLLM Leaderboard (scores.json) Official board, 16 Apr 2026 |
Tax Agent Bench 2026-10-29
accuracy % (all-pass x citation factor), higher is better. Top 5 of 10 models.
Updated 29 Oct 2026
- 1Qwen 3.8 Max 090249.7
- 2Claude Fable 5.149.2
- 3Gemini 4 Carbon48.9
- 4Claude Opus 5.545.6
- 5Claude Opus 545.5
| 1 | Qwen 3.8 Max 0902 | Qwen | 49.7 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
|---|---|---|---|---|---|
| 2 | Claude Fable 5.1 | Anthropic | 49.2 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 3 | Gemini 4 Carbon | 48.9 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 | |
| 4 | Claude Opus 5.5 | Anthropic | 45.6 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 5 | Claude Opus 5 | Anthropic | 45.5 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 6 | Gemini 4 Argon | 44.6 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 | |
| 7 | DeepSeek V4.1 Flash 1017 | DeepSeek | 44.4 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 8 | Wonder Full | Not stated | 43 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 9 | Muse Spark 1.3 Max | Not stated | 42.4 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
| 10 | Claude Sonnet 5.5 | Anthropic | 42.0 | test set | Vals.ai leaderboard Official board, 29 Oct 2026 |
Advanced Financial Reasoning at Scale (CFA Level III evaluation) arXiv v2 (2025-09-22); FinLLM @ IJCAI 2025
overall % (mean of MCQ and essay), higher is better. Top 5 of 8 models.
Updated 22 Sept 2025
- 1o4-mini79.1
- 2Gemini 2.5 Flash77.3
- 3Gemini 2.5 Pro75.9
- 4Claude Opus 474.9
- 5o3-mini74.4
| 1 | o4-mini | OpenAI | 79.1 | MCQ Self-Discover; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
|---|---|---|---|---|---|
| 2 | Gemini 2.5 Flash | 77.3 | MCQ zero-shot; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 | |
| 3 | Gemini 2.5 Pro | 75.9 | MCQ zero-shot; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 | |
| 4 | Claude Opus 4 | Anthropic | 74.9 | MCQ CoT-SC N=3; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
| 5 | o3-mini | OpenAI | 74.4 | MCQ CoT-SC N=5; essay Self-Discover or CoT-SC N=5 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
| 6 | Grok-3 | xAI | 73.8 | MCQ CoT-SC N=3; essay CoT-SC N=5 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
| 7 | Claude Sonnet 4 | Anthropic | 73.4 | MCQ CoT-SC N=3; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
| 8 | Grok-3 Mini Low Effort | xAI | 72.4 | MCQ CoT-SC N=3; essay CoT-SC N=3 | Paper Table 1 Reported by the authors, 22 Sept 2025 |
Reasoning Models Ace the CFA Exams (CFA mock-exam evaluation) arXiv v1 (2025-12-09)
Level I accuracy %, higher is better. Top 5 of 6 models.
Updated 9 Dec 2025
- 1Gemini 3.0 Pro97.6
- 2gpt-596.1
- 3Gemini 2.5 Pro95.7
- 4grok-494.8
- 5Claude Opus 4.194.6
| 1 | Gemini 3.0 Pro | 97.6 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 | |
|---|---|---|---|---|---|
| 2 | gpt-5 | OpenAI | 96.1 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 |
| 3 | Gemini 2.5 Pro | 95.7 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 | |
| 4 | grok-4 | xAI | 94.8 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 |
| 5 | Claude Opus 4.1 | Anthropic | 94.6 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 |
| 6 | DeepSeek V3.1 | DeepSeek | 90.9 | zero-shot | Paper Table 5 Reported by the authors, 9 Dec 2025 |