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 1FB Maintainedverificationforensic

Pave Interchange Fee Benchmark

Forensic financial verification across 50,000 synthetic transaction records spanning four complexity levels from deterministic aggregation to strategic impact modelling.

24 tasks

8 models evaluated

Updated 2026

FinanceBenchmark (2026)

View Source →Submit your score Methodology

100.0%

Top Score

Top model

avae-2-0

Model Rankings

Overall score across all four complexity levels, weighted equally.

financebenchmark.ai · Pave

66.0%

🥇

AnthropicClaude Opus 4.773.0%

🥈

OpenAIGPT-5.571.0%

🥉

GoogleGemini 3.1 Pro69.0%

4

DeepSeekDeepSeek V4 ProOPEN65.0%

5

xAIGrok 4.368.0%

6

Moonshot AIKimi K2.6OPEN66.0%

7

Zhipu AIGLM-5.1OPEN63.0%

8

MiniMaxMiniMax M2.7OPEN62.0%

Full Results Table

#	Model	Overall	Level 1	Level 2	Level 3	Level 4	Source	Date
🥇	Claude Opus 4.7 Anthropic	73.0	96.0	75.0	52.0	68.0	FinanceBenchmark (2026)	2026-05
🥈	GPT-5.5 OpenAI	71.0	95.0	72.0	50.0	66.0	FinanceBenchmark (2026)	2026-05
🥉	Gemini 3.1 Pro Google	69.0	94.0	70.0	48.0	65.0	FinanceBenchmark (2026)	2026-05
4	Grok 4.3 xAI	68.0	93.0	68.0	47.0	63.0	FinanceBenchmark (2026)	2026-05
5	Kimi K2.6OPEN Moonshot AI	66.0	92.0	66.0	45.0	61.0	FinanceBenchmark (2026)	2026-05
6	DeepSeek V4 ProOPEN DeepSeek	65.0	91.0	65.0	44.0	60.0	FinanceBenchmark (2026)	2026-05
7	GLM-5.1OPEN Zhipu AI	63.0	90.0	63.0	42.0	58.0	FinanceBenchmark (2026)	2026-05
8	MiniMax M2.7OPEN MiniMax	62.0	89.0	61.0	40.0	56.0	FinanceBenchmark (2026)	2026-05