OmniaBench - ToB: leaderboard

Metric: Pass@1 (%; tasks from the business-facing (ToB) domain split; a task counts as solved only when every rubric item is satisfied (Program tasks by a VerifyCode check); up to 200 tool-call steps; challenging set of 644 tasks evaluated June 29 to July 15, 2026). Source: arxiv.org. Saturation forecast: Around June 2027. 22 models tracked.

Top models

#ModelScore
1Claude Sonnet 5 (Max)60.68
2GPT-5.5 (High)58.2
3GPT-5.6 Sol (High)57.59
4GLM-5.2 (Max)57.28
5Claude Opus 4.7 (Max)56.66
6Kimi K2.6 (Thinking)54.8
7DeepSeek V4 Pro (Max)54.18
8Qwen 3.7 Max (Thinking)48.3
9GLM-5 (Thinking)47.68
10Gemini 3.5 Flash (High)46.75
11DeepSeek V4 Flash (Max)45.2
12Qwen 3.5 397B A17B44.27
13Qwen 3.6 27B43.65
14Gemini 3.1 Pro (Preview)42.11
15Qwen 3.6 35B A3B41.8

Interactive version: theaggregate.ai/benchmark?slug=omniabench-tob · How It Works · Data refreshed daily, snapshot 2026-09-29.