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

Terminal-Bench 2.1

Interactive Linux terminal command execution, shell debugging, and agent environment handling.

Variable tasks

23 models evaluated

Updated 2026

View Source →Submit your score Methodology

85.8%

Top Score

Top model

GPT-5.6 Sol

Model Rankings

financebenchmark.ai · Terminal-Bench

80.9%

🥇

AnthropicClaude Opus 584.6%

🥈

AnthropicClaude Sonnet 574.5%

🥉

OpenAIGPT-5.6 Sol85.8%

4

DeepSeekDeepSeek V4 FlashOPEN67.0%

5

xAIGrok 4.678.3%

6

Moonshot AIKimi K3OPEN80.9%

7

AnthropicClaude Fable 580.5%

8

DeepSeekDeepSeek V4 Pro (0813)OPEN54.7%

9

GoogleGemini 3.5 Flash Lite50.2%

10

GoogleGemini 3.6 Flash66.8%

11

Zhipu AIGLM 5.2OPEN67.8%

12

OpenAIGPT-5.6 Luna79.0%

13

OpenAIGPT-5.6 Terra73.4%

14

xAIGrok 4.567.8%

15

Thinking MachinesInklingOPEN47.6%

16

Thinking MachinesInkling SmallOPEN55.0%

17

Moonshot AIKimi K2.7 CodeOPEN67.0%

18

Ant GroupLing 3.0 FlashOPEN50.2%

19

MetaMuse Spark 1.169.3%

20

MetaMuse Spark 1.269.7%

21

NVIDIANemotron 3 UltraOPEN50.9%

22

AlibabaQwen 3.7 Plus52.8%

23

AlibabaQwen 3.8 MaxOPEN67.2%

Full Results Table

#	Model	Overall	Source	Date
🥇	GPT-5.6 Sol OpenAI	85.8	VALS (2026)	2026-07
🥈	Claude Opus 5 Anthropic	84.6	VALS (2026)	2026-07
🥉	Kimi K3OPEN Moonshot AI	80.9	VALS (2026)	2026-07
4	Claude Fable 5 Anthropic	80.5	VALS (2026)	2026-06
5	GPT-5.6 Luna OpenAI	79.0	VALS (2026)	2026-07
6	Grok 4.6 xAI	78.3	VALS (2026)	2026-08
7	Claude Sonnet 5 Anthropic	74.5	VALS (2026)	2026-06
8	GPT-5.6 Terra OpenAI	73.4	VALS (2026)	2026-07
9	Muse Spark 1.2 Meta	69.7	VALS (2026)	2026-08
10	Muse Spark 1.1 Meta	69.3	VALS (2026)	2026-07
11	GLM 5.2OPEN Zhipu AI	67.8	VALS (2026)	2026-06
12	Grok 4.5 xAI	67.8	VALS (2026)	2026-07
13	Qwen 3.8 MaxOPEN Alibaba	67.2	VALS (2026)	2026-08
14	DeepSeek V4 FlashOPEN DeepSeek	67.0	VALS (2026)	2026-07
15	Kimi K2.7 CodeOPEN Moonshot AI	67.0	VALS (2026)	2026-06
16	Gemini 3.6 Flash Google	66.8	VALS (2026)	2026-07
17	Inkling SmallOPEN Thinking Machines	55.0	VALS (2026)	2026-07
18	DeepSeek V4 Pro (0813)OPEN DeepSeek	54.7	VALS (2026)	2026-08
19	Qwen 3.7 Plus Alibaba	52.8	VALS (2026)	2026-06
20	Nemotron 3 UltraOPEN NVIDIA	50.9	VALS (2026)	2026-06
21	Gemini 3.5 Flash Lite Google	50.2	VALS (2026)	2026-07
22	Ling 3.0 FlashOPEN Ant Group	50.2	VALS (2026)	2026-07
23	InklingOPEN Thinking Machines	47.6	VALS (2026)	2026-07