DAYJOB: Accounting: leaderboard
Long-horizon accounting assignments from a brief request and a workspace of files, graded by an agent judge against an expert rubric; the score is the share of attempts meeting at least 95% of criteria. Surge AI supplied these results before publishing the board and has not stated the task count.
Metric: Attempts meeting at least 95% of rubric criteria (%). Source: fireworks.ai. Saturation forecast: Around February 2028. 16 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Sonnet 5.5 | Max | 35.2 | #19 |
| 2 | Claude Opus 5.5 | Max | 34.4 | #5 |
| 3 | Muse Wonder | Max | 30.8 | #28 |
| 4 | GPT-6 Astra | Max | 30 | #6 |
| 5 | Claude Fable 5.1 | Max | 27.2 | #15 |
| 6 | GPT-6.1 Sol | Max | 25.2 | #10 |
| 7 | Grok 4.7 | xHigh | 20 | #47 |
| 8 | Gemini 4 Argon | High | 18 | #12 |
| 9 | Muse Spark 1.3 | Max | 18 | #43 |
| 10 | Kimi K3.1 Pro | 13.6 | #29 | |
| 11 | Hy4 preview | High | 8.4 | #71 |
| 12 | GLM-5.3 | Max | 7.6 | #68 |
| 13 | DeepSeek V4.1 Flash (1017) | Max | 7.6 | #41 |
| 14 | Kimi K3 | Max | 6 | #56 |
| 15 | Gemini 3.8 Flash | High | 6 | #36 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon.
Interactive version: theaggregate.ai/benchmark?slug=dayjob-accounting · How It Works · Data refreshed daily, snapshot 2026-11-16.