Deep FinResearch Bench

Size

100 professional reports covering 25 S&P 500 companies (FY2025 Q1-Q2), each paired with an AI-generated report

Public split

none (the paper links no data or code; the professional reports came from a paid Yahoo Finance subscription)

Grading

Automated. GPT-5 extracts claims, GPT-4.1 checks each against the cited page, and each claim ends up factual, hallucinated or non-verifiable. A GPT-5 judge gives 1-4 rubric scores, and forecasts are scored on SMAPE, hit rate and directional accuracy.

Licence

Not stated

Submission

none

Products allowed

Yes

Latest version

arXiv v1 (2026-04-22)

What it does not cover

No public data or harness. Covers only 25 large-cap US names in IT, financials and health care, over two quarters.

Each score shows who reported it: the benchmark's authors (quoted from their paper), the benchmark's official board, or Moneda (run with the benchmark's own grader and rules). We do not copy other aggregators.

1. 86 OpenAI86
2. 75.6Perplexity75.6
3. 69.6Gemini69.6
4. 53.2 Grok53.2

1. 1 OpenAI8686
2. 2Perplexity75.675.6
3. 3Gemini69.669.6
4. 4 Grok53.253.2

Deep FinResearch Bench leaderboard (overall factual claim rate %, higher is better)

Rank	Model	Organisation	overall factual claim rate %	Setting	Source
1	OpenAI	OpenAI	86	OpenAI DR (GPT-o3), Firm A-structured reports	Paper Table 7 Reported by the authors, 22 Apr 2026
2	Perplexity	Perplexity	75.6	Perplexity DR (Sonar Reasoning), Firm A-structured reports	Paper Table 7 Reported by the authors, 22 Apr 2026
3	Gemini	Google	69.6	Gemini DR (Gemini 2.5 Pro), Firm A-structured reports	Paper Table 7 Reported by the authors, 22 Apr 2026
4	Grok	xAI	53.2	Grok DR (Grok 4), Firm A-structured reports	Paper Table 7 Reported by the authors, 22 Apr 2026

1. 2.3Gemini2.3
2. 2.1Perplexity2.1
3. 2.0 Grok2.0
4. 2.0 OpenAI2.0

1. 1Gemini2.32.3
2. 2Perplexity2.12.1
3. 3 Grok2.02.0
4. 3 OpenAI2.02.0

Deep FinResearch Bench leaderboard (qualitative overall (1-4, GPT-5 judge), higher is better)

Rank	Model	Organisation	qualitative overall (1-4, GPT-5 judge)	Setting	Source
1	Gemini	Google	2.3	Default	Paper Table 3 Reported by the authors, 22 Apr 2026
2	Perplexity	Perplexity	2.1	Default	Paper Table 3 Reported by the authors, 22 Apr 2026
3	Grok	xAI	2.0	Default	Paper Table 3 Reported by the authors, 22 Apr 2026
3	OpenAI	OpenAI	2.0	Default	Paper Table 3 Reported by the authors, 22 Apr 2026

• Paper

Last checked 1 Nov 2026. Something out of date or wrong? Send a correction.