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 3agentTasks

LiveCodeBench

Contamination-free competitive programming and code completion benchmark.

Variable tasks

19 models evaluated

Updated 2026

arXiv:2403.07974

View Source →Submit your score Methodology

89.8%

Top Score

Top model

Claude Fable 5

Model Rankings

financebenchmark.ai · LiveCodeBench

87.2%

🥇

AnthropicClaude Opus 589.0%

🥈

AnthropicClaude Sonnet 582.4%

🥉

OpenAIGPT-5.6 Sol82.6%

4

DeepSeekDeepSeek V4 FlashOPEN87.3%

5

xAIGrok 4.688.2%

6

Moonshot AIKimi K3OPEN87.2%

7

AnthropicClaude Fable 589.8%

8

GoogleGemini 3.5 Flash Lite79.0%

9

GoogleGemini 3.6 Flash88.0%

10

Zhipu AIGLM 5.2OPEN69.5%

11

OpenAIGPT-5.6 Terra85.9%

12

xAIGrok 4.587.3%

13

Thinking MachinesInklingOPEN85.5%

14

Thinking MachinesInkling SmallOPEN85.9%

15

Moonshot AIKimi K2.7 CodeOPEN82.0%

16

Ant GroupLing 3.0 FlashOPEN84.0%

17

MetaMuse Spark 1.185.9%

18

NVIDIANemotron 3 UltraOPEN86.0%

19

AlibabaQwen 3.8 MaxOPEN87.8%

Full Results Table

#	Model	Overall	Source	Date
🥇	Claude Fable 5 Anthropic	89.8	arXiv:2403.07974	2026-06
🥈	Claude Opus 5 Anthropic	89.0	arXiv:2403.07974	2026-07
🥉	Grok 4.6 xAI	88.2	arXiv:2403.07974	2026-08
4	Gemini 3.6 Flash Google	88.0	arXiv:2403.07974	2026-07
5	Qwen 3.8 MaxOPEN Alibaba	87.8	arXiv:2403.07974	2026-08
6	Grok 4.5 xAI	87.3	arXiv:2403.07974	2026-07
7	DeepSeek V4 FlashOPEN DeepSeek	87.3	arXiv:2403.07974	2026-07
8	Kimi K3OPEN Moonshot AI	87.2	arXiv:2403.07974	2026-07
9	Nemotron 3 UltraOPEN NVIDIA	86.0	arXiv:2403.07974	2026-06
10	GPT-5.6 Terra OpenAI	85.9	arXiv:2403.07974	2026-07
11	Inkling SmallOPEN Thinking Machines	85.9	arXiv:2403.07974	2026-07
12	Muse Spark 1.1 Meta	85.9	arXiv:2403.07974	2026-07
13	InklingOPEN Thinking Machines	85.5	arXiv:2403.07974	2026-07
14	Ling 3.0 FlashOPEN Ant Group	84.0	arXiv:2403.07974	2026-07
15	GPT-5.6 Sol OpenAI	82.6	arXiv:2403.07974	2026-07
16	Claude Sonnet 5 Anthropic	82.4	arXiv:2403.07974	2026-06
17	Kimi K2.7 CodeOPEN Moonshot AI	82.0	arXiv:2403.07974	2026-06
18	Gemini 3.5 Flash Lite Google	79.0	arXiv:2403.07974	2026-07
19	GLM 5.2OPEN Zhipu AI	69.5	arXiv:2403.07974	2026-06