CLBench Life - Group Conversations & Meeting Transcripts: leaderboard

CL-bench Life (2026) tests whether language models can learn from long, messy real-life context to solve human-written tasks, a task counting as solved only when every expert rubric item is met. This board is the task solving rate in percent on tasks over group chats and meeting transcripts; higher is better.

Metric: Task solving rate (%) on the 45 group-conversation and meeting-transcript tasks of CL-bench Life; a task is solved only when the response satisfies every human-written binary rubric item, judged by GPT-5.1 at high reasoning effort from the response and rubrics alone; reasoning mode at high effort where the model supports it, mean of three runs; higher is better. Source: arxiv.org. Saturation forecast: Around March 2028. 10 models tracked. Updated 7 Oct 2026.

Top models

#ModelSettingScoreOverall rank
1GPT-5.4High30.4#138
2Gemini 3.1 Pro (Preview)High28.9#102
3Claude Opus 4.6High28.1#114
4Seed 2.0 ProHigh26.7#226
5Kimi K2.5High23.7#272
6Hy3-previewHigh22.2#320
7DeepSeek V3.2Thinking17#425
8Grok 4.20High16.3#307
9Qwen 3.5 PlusHigh14.8#250
10MiniMax-M2.5Thinking10.4#600

No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.

Interactive version: theaggregate.ai/benchmark?slug=clbench-life-group-conversations-meeting-transcripts · How It Works · Data refreshed daily, snapshot 2026-11-22.