Atlas

Benchmarks

← All benchmarks

DAYJOB: Finance (October 2026 rescoring)

Finance & Accounting

Surge's DAYJOB: Finance board as rescored between its October 23 and October 27, 2026 site builds, on 80 assignments built by finance professionals. The page has no changelog and names no metric or rubric threshold. The paper and dataset card still describe passing only on every rubric criterion, and the launch post still quotes the old 23.9% leader. Surge doesn't say what the new scores measure, but against the paper's strict column they reproduce its statistics for a 95% rubric threshold, which Surge states only for DAYJOB: Tax: Kendall tau-b 0.92 and the top five multiplied by 1.8 on average. Every value fits 400 attempts, five per task as the linked harness repository states; that repository, unchanged since September 23, also names a Claude Opus 4.8 judge. Kept apart from the original board because scores rose by as much as 18.6 points (Claude Opus 5.5 Max, 23.9 to 42.5).

Benchmark source: Surgehq

Top models (higher is better)

ModelScoreSource
Claude Opus 5.542.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Claude Sonnet 5.540.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Muse Wonder38.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-6 Astra37.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Claude Fable 5.136.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-6.1 Sol33.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Gemini 4 Argon30.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Muse Spark 1.327.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Grok 4.726.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Claude Opus 523.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Grok 4.621.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-6 Sol19.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Claude Fable 518.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
DeepSeek V4.1 Flash 101712.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-5.6 Sol11.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GLM-5.311.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GLM-5.3 Flash7.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-6 Luna6.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Muse Spark 1.26.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Kimi K35.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Gemini 3.7 Flash5.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Mistral Large 4 Preview4.8Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Sonnet 54.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Gemini 3.8 Flash4.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-5.6 Luna4.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
GPT-5.6 Terra3.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Beam 1.0 501B A23B1.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Hy30.5Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Gemini 3.1 Pro Preview0.3Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Inkling0.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Kimi K2.7 Code0.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Mistral Large 3 675B Instruct 25120.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Muse Glimmer 30B0.0Surge AI DAYJOB: Finance (October 2026 rescoring) leaderboard · Surge AI
Loading Atlas data…