CompanyBench
Agents complete financial analyst workflows in a simulated bank. Checks compare outputs and database changes with expected results.
Each model attempted each task ten times.
- Made by
- Turing
- Kind
- Finance
- Level
- Expert
- Scored
- Checked automatically
- Answers
- Agent task
- Tasks
- Kept private
- Measures
- Capability
- Subject
- Economics and business
- Interaction
- Agentic
- Environment
- A simulator
- Refresh
- One fixed set
- Run by
- A host
- Occupations
- Business and financial operations
- In views
- Agents, Robotics, Law and finance
- Strongest model tested
- GPT-5.5 #21 on the capability index
- Last model added
- May 2026
- Scores refreshed
- November 20th, 2026
GPT-5.5 leads with 44.4% among models released by May 2026
Best score per model from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.
The newest model with a score here came out in May 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:
- GPT-6.2 Astra
- Claude Fable 5.2
-
Gemini 4 Carbon
- GPT-6 Astra
- and 16 more
- 1 GPT-5.5 44.4%
- 2 Claude Opus 4.8 35.9%
- 3
Gemini 3.1 Pro 18.3%
- 4
Gemini 3.5 Flash 7.66%
Source: Official CompanyBench leaderboard
| # | Model | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | GPT-5.5 | 44.4% | Apr 2026 | Official leaderboard | |
| 2 | Claude Opus 4.8 | 35.9% | May 2026 | Official leaderboard | |
| 3 | | 18.3% | Feb 2026 | Official leaderboard | |
| 4 | | 7.66% | May 2026 | Official leaderboard |
Rankings by part
We read 9 parts from the benchmark’s own leaderboard, each ranking models on one part of the test, such as Reconciliation & Settlement, Deposits & Funding and Customer & Account Analytics. GPT-5.5 leads 7 of them, and Claude Opus 4.8 leads the other 2. Arrows show how far a model moved from its overall rank among the models we track.
Reconciliation & Settlement
4 models we track have a score here.
- 1 Claude Opus 4.8 ↑1 30%
- 2 GPT-5.5 ↓1 23.3%
- 3
Gemini 3.1 Pro 5.00%
- 4
Gemini 3.5 Flash 3.30%
Deposits & Funding
4 models we track have a score here.
- 1 GPT-5.5 55%
- 2 Claude Opus 4.8 10%
- 3
Gemini 3.1 Pro 0%
- 4
Gemini 3.5 Flash 0%
Customer & Account Analytics
4 models we track have a score here.
- 1 GPT-5.5 40%
- 2 Claude Opus 4.8 34%
- 3
Gemini 3.1 Pro 2.00%
- 4
Gemini 3.5 Flash 0%
Revenue & Fees
4 models we track have a score here.
- 1 GPT-5.5 33.3%
- 2 Claude Opus 4.8 25%
- 3
Gemini 3.1 Pro 18.3%
- 4
Gemini 3.5 Flash 5.00%
Compliance & Regulatory
4 models we track have a score here.
- 1 GPT-5.5 44.3%
- 2 Claude Opus 4.8 25.7%
- 3
Gemini 3.1 Pro 24.3%
- 4
Gemini 3.5 Flash 2.90%
Cards & Interchange
4 models we track have a score here.
- 1 Claude Opus 4.8 ↑1 52.5%
- 2 GPT-5.5 ↓1 43.3%
- 3
Gemini 3.1 Pro 24.2%
- 4
Gemini 3.5 Flash 2.50%
Payments & Transfers
4 models we track have a score here.
- 1 GPT-5.5 56%
- 2 Claude Opus 4.8 43%
- 3
Gemini 3.1 Pro 22%
- 4
Gemini 3.5 Flash 2.00%
Risk & Fraud
4 models we track have a score here.
- 1 GPT-5.5 50.8%
- 2 Claude Opus 4.8 35.8%
- 3
Gemini 3.5 Flash ↑1 22.5%
- 4
Gemini 3.1 Pro ↓1 14.2%
Subscriptions & Billing
4 models we track have a score here.
- 1 GPT-5.5 47.5%
- 2
Gemini 3.1 Pro ↑1 42.5%
- 3 Claude Opus 4.8 ↓1 27.5%
- 4
Gemini 3.5 Flash 25%
2 models have held number one on the official leaderboard since 2023
Lab claims count from the day the card came out, independent runs from the model’s release.
Source: Official CompanyBench leaderboard
Scores over time
All the scores we have, one dot each, so a model tested at several settings appears more than once.
Source: Official CompanyBench leaderboard
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Law, finance and medicine
- MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy.
Qwen 3.8 Max 80.4%
- HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
- HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
- FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy.
Ling-3.1-flash 81.7%
- KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
- Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%
Data audit
8 of 9 checks pass, 1 warning. Look first at the Newest score check. The newest came 6 months ago, in May 2026, for Claude Opus 4.8. Every benchmark
Catalog record 2 pass
- Pass Task access The tasks are private.
- Pass Confidence Checked against a primary source.
Scores 1 warning 2 pass 2 notes
- Pass Independent runs Scores from the official leaderboard (4 entries).
- Note Models scored Only 4 models we track.
- Warning Newest score The newest came 6 months ago, in May 2026, for Claude Opus 4.8.
- Note Frontier models None of the top 10 has a score. The strongest tested is GPT-5.5 (#21).
- Pass Headroom The best is 44.4%, under 85%, so it still separates models.
Leaderboard import 3 pass
- Pass Importer We read the board on bench.turing.com: 4 rows for models we track at the last read.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.