ATLAS-Finance: leaderboard
An agent inside a simulated advisory firm finds its assignment in inboxes, chat and drives, builds a live-formula Excel model and sends it to the right people; the score is the pass rate over 3 runs of 100 tasks, run by Handshake in the OpenCode harness. Inputs are scattered and sometimes corrected mid-task, unlike single-prompt finance tests.
Metric: Tasks passed, mean of 3 runs of 100 tasks (%). Source: fireworks.ai. Saturation forecast: Around 2029. 11 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 5 | High | 12.33 | #20 |
| 2 | Claude Fable 5.1 | High | 12 | #15 |
| 3 | Claude Opus 5.5 | High | 11.41 | #5 |
| 4 | GPT-6 Astra | High | 11.33 | #6 |
| 5 | Grok 4.7 | High | 5 | #47 |
| 6 | Gemini 3.8 Flash | High | 4.67 | #36 |
| 7 | GPT-5.6 Sol | High | 4.67 | #30 |
| 8 | Kimi K3 | High | 4.67 | #56 |
| 9 | GLM-5.3 | High | 4.67 | #68 |
| 10 | DeepSeek V4 Pro | High | 1.67 | #191 |
| 11 | GPT-5.6 Luna | High | 0.33 | #122 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #10 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=atlas-finance · How It Works · Data refreshed daily, snapshot 2026-11-16.