OmniBehavior - Ads Binary Behaviors: leaderboard

OmniBehavior (2026) tests LLMs as simulators of real users: given the profile and long, real behaviour history of a user across the Kuaishou video, live-streaming, ads and e-commerce scenarios, the model predicts what the user does next. F1 on binary engagement decisions in the advertising scenario, in percent; higher is better.

Metric: F1 (%) on binary engagement decisions in the advertising scenario, for predicting the real behaviour of 200 Kuaishou users (three months of interleaved video, live-stream, ads and e-commerce traces, 8,143 actions per user on average) from the user profile, history and current scenario, 6,000 balanced tasks, 32k context window, temperature 0.1; higher is better. Source: arxiv.org. Saturation forecast: Around February 2027. 11 models tracked. Updated 7 Oct 2026.

Top models

#ModelScoreOverall rank
1Claude Opus 4.551.16#160
2Claude Sonnet 4.542.77#261
3GLM-4.740.34#380
4Claude Sonnet 436.81#410
5GPT-5.233.6#185
6Kimi K2 090531.19#570
7Claude Haiku 4.530#583
8OmniBehavior DeepSeek-V3 (checkpoint unspecified)25.74
9GPT-4o25.24#751
10Gemini 3 Flash24.64#175
11Qwen 3 235B A22B 2507 Instruct23.19#616

No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=omnibehavior-ads-binary-behaviors · How It Works · Data refreshed daily, snapshot 2026-11-21.