CarryOnBench - Utility (Benign Clarifications): leaderboard
CarryOnBench (2026) starts from seemingly harmful requests that have a benign purpose and checks whether a model gives the benign information a user needs, and how harmful its answers get, as simulated users clarify or push back. Share of the benign information checklist fulfilled in multi-turn conversations where the user clarifies the benign intent, counted over the whole conversation, in percent; higher is better.
Metric: Ben-Util (%): share of the benign information-need checklist (about 18 atomic yes/no items per query, built from five models' compliant answers) that the responses fulfil, each item judged separately by Gemini-2.5-Flash, over multi-turn conversations of 4 to 12 turns in which at least one user turn clarifies the benign intent, conversation-level, on CarryOnBench (398 seemingly harmful SORRY-Bench queries with a benign underlying intent; user turns simulated by Gemini-3-Flash); higher is better. Source: arxiv.org. Saturation forecast: Not forecast. 14 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 3 Next 80B A3B Instruct | 75.2 | #778 |
| 2 | GPT-5 Mini | 74.8 | #364 |
| 3 | Qwen 3 32B | 73.8 | #972 |
| 4 | Qwen 3 235B A22B | 71.2 | #628 |
| 5 | DeepSeek V3.1 | 66.3 | #511 |
| 6 | GPT-5.4 | 65.4 | #135 |
| 7 | OLMo 3.1 32B Instruct | 64.4 | #2445 |
| 8 | OLMo 3 7B Instruct | 61.3 | #6953 |
| 9 | Gemini 3.1 Pro (Preview) | 57.4 | #99 |
| 10 | GPT-OSS-120B | 52.5 | #704 |
| 11 | Claude Sonnet 4 | 51.3 | #410 |
| 12 | Claude Haiku 4.5 | 49.9 | #583 |
| 13 | Mixtral 8x7B Instruct | 45.1 | #4696 |
| 14 | Llama 4 Maverick | 37 | #1110 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=carryonbench-utility-benign-clarifications · How It Works · Data refreshed daily, snapshot 2026-11-21.