FinanceBenchmark: Financial AI Leaderboard 2026

FinanceBenchmark provides independent rankings and a composite score for leading Artificial Intelligence models, including Anthropic Claude, OpenAI GPT, and Google Gemini, on complex financial reasoning and forensic accounting tasks. Enable JavaScript to view the interactive, continuously updated leaderboard.

2026-09-10MODELDeepSeek-V4.1-Flash released with extreme agent memory efficiency and evaluated across VALS • 2026-09-03MODELGPT-6 Astra released by OpenAI - 71.7% on EMB, 63.3% on Tax Agent Bench • 2026-09-02MODELGoogle Gemini 3.8 Flash takes #1 rank on Finance Agent (v2) with 61.4% and 72.2% on EMB • 2026-09-02MODELMeta releases Muse Spark 1.3 & 1.3 Max - 59.0%-60.0% on Finance Agent (v2) • 2026-09-01MODELClaude Fable 5.1 leads Tax Agent Bench (77.6%) and EMB (76.7%) with 75% cheaper cache reads • 2026-09-11BENCHTax Agent Bench and FAB v2 benchmarks updated with latest frontier model rankings • 2026-09-15UPDATEPlatform updated: 43 models tracked across 41 independent benchmarks • 2026-09-10MODELDeepSeek-V4.1-Flash released with extreme agent memory efficiency and evaluated across VALS • 2026-09-03MODELGPT-6 Astra released by OpenAI - 71.7% on EMB, 63.3% on Tax Agent Bench • 2026-09-02MODELGoogle Gemini 3.8 Flash takes #1 rank on Finance Agent (v2) with 61.4% and 72.2% on EMB • 2026-09-02MODELMeta releases Muse Spark 1.3 & 1.3 Max - 59.0%-60.0% on Finance Agent (v2) • 2026-09-01MODELClaude Fable 5.1 leads Tax Agent Bench (77.6%) and EMB (76.7%) with 75% cheaper cache reads • 2026-09-11BENCHTax Agent Bench and FAB v2 benchmarks updated with latest frontier model rankings • 2026-09-15UPDATEPlatform updated: 43 models tracked across 41 independent benchmarks •

Tier 2numericalverification

GPQA Diamond

Graduate-level Google-proof Q&A benchmark across physics, chemistry, biology, and math.

198 tasks

20 models evaluated

Updated 2026

arXiv:2311.12022

View Source →Submit your score Methodology

95.2%

Top Score

Top model

GPT-5.6 Sol

Model Rankings

financebenchmark.ai · GPQA Diamond

92.9%

🥇

AnthropicClaude Opus 593.4%

🥈

AnthropicClaude Sonnet 588.9%

🥉

OpenAIGPT-5.6 Sol95.2%

4

DeepSeekDeepSeek V4 FlashOPEN89.9%

5

xAIGrok 4.694.7%

6

Moonshot AIKimi K3OPEN92.9%

7

AnthropicClaude Fable 593.2%

8

DeepSeekDeepSeek V4 Pro (0813)OPEN92.4%

9

GoogleGemini 3.5 Flash Lite83.8%

10

GoogleGemini 3.6 Flash93.4%

11

Zhipu AIGLM 5.2OPEN85.6%

12

OpenAIGPT-5.6 Luna91.7%

13

OpenAIGPT-5.6 Terra90.9%

14

xAIGrok 4.592.9%

15

Thinking MachinesInklingOPEN87.1%

16

Thinking MachinesInkling SmallOPEN83.6%

17

Ant GroupLing 3.0 FlashOPEN84.8%

18

MetaMuse Spark 1.191.2%

19

NVIDIANemotron 3 UltraOPEN86.1%

20

AlibabaQwen 3.8 MaxOPEN93.7%

Full Results Table

#	Model	Overall	Source	Date
🥇	GPT-5.6 Sol OpenAI	95.2	arXiv:2311.12022	2026-07
🥈	Grok 4.6 xAI	94.7	arXiv:2311.12022	2026-08
🥉	Qwen 3.8 MaxOPEN Alibaba	93.7	arXiv:2311.12022	2026-08
4	Claude Opus 5 Anthropic	93.4	arXiv:2311.12022	2026-07
5	Gemini 3.6 Flash Google	93.4	arXiv:2311.12022	2026-07
6	Claude Fable 5 Anthropic	93.2	arXiv:2311.12022	2026-06
7	Kimi K3OPEN Moonshot AI	92.9	arXiv:2311.12022	2026-07
8	Grok 4.5 xAI	92.9	arXiv:2311.12022	2026-07
9	DeepSeek V4 Pro (0813)OPEN DeepSeek	92.4	arXiv:2311.12022	2026-08
10	GPT-5.6 Luna OpenAI	91.7	arXiv:2311.12022	2026-07
11	Muse Spark 1.1 Meta	91.2	arXiv:2311.12022	2026-07
12	GPT-5.6 Terra OpenAI	90.9	arXiv:2311.12022	2026-07
13	DeepSeek V4 FlashOPEN DeepSeek	89.9	arXiv:2311.12022	2026-07
14	Claude Sonnet 5 Anthropic	88.9	arXiv:2311.12022	2026-06
15	InklingOPEN Thinking Machines	87.1	arXiv:2311.12022	2026-07
16	Nemotron 3 UltraOPEN NVIDIA	86.1	arXiv:2311.12022	2026-06
17	GLM 5.2OPEN Zhipu AI	85.6	arXiv:2311.12022	2026-06
18	Ling 3.0 FlashOPEN Ant Group	84.8	arXiv:2311.12022	2026-07
19	Gemini 3.5 Flash Lite Google	83.8	arXiv:2311.12022	2026-07
20	Inkling SmallOPEN Thinking Machines	83.6	arXiv:2311.12022	2026-07