Hebbia Financial Services Benchmark: leaderboard
Analyst requests answered over document sets, as Excel models or as client-ready decks, run by Hebbia in its production agent harness; the score is the strict pass rate on 151 tasks. The deliverables are files a client would receive, not short answers.
Metric: Strict pass rate on 151 tasks (%). Source: fireworks.ai. Saturation forecast: Around May 2027. 10 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 5.5 | Max | 74.6 | #5 |
| 2 | GPT-6.1 Sol | Max | 73.5 | #10 |
| 3 | GPT-6 Astra | Max | 72.3 | #6 |
| 4 | DeepSeek V4.1 Flash | Max | 63.1 | #62 |
| 5 | Gemini 3.8 Flash | 60.9 | #36 | |
| 6 | Kimi K3 | Max | 60.3 | #56 |
| 7 | DeepSeek V4.1 Flash (1017) | Max | 60.3 | #41 |
| 8 | Muse Spark Open | Max | 49 | #81 |
| 9 | MiniMax-M3.1-Flash | Max | 47 | #109 |
| 10 | GLM-5.3 Flash | Max | 43.7 | #125 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=hebbia-financial-services-benchmark · How It Works · Data refreshed daily, snapshot 2026-11-16.