CarryOnBench - Utility (Benign Clarifications): leaderboard
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.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 Next 80B A3B Instruct | 75.2 |
| 2 | GPT-5 Mini | 74.8 |
| 3 | Qwen 3 32B | 73.8 |
| 4 | Qwen 3 235B A22B | 71.2 |
| 5 | DeepSeek V3.1 | 66.3 |
| 6 | GPT-5.4 | 65.4 |
| 7 | Olmo 3.1 32B Instruct | 64.4 |
| 8 | OLMo 3 7B Instruct | 61.3 |
| 9 | Gemini 3.1 Pro (Preview) | 57.4 |
| 10 | GPT-OSS-120B | 52.5 |
| 11 | Claude Sonnet 4 | 51.3 |
| 12 | Claude Haiku 4.5 | 49.9 |
| 13 | Mixtral 8x7B Instruct (v0.1) | 45.1 |
| 14 | Llama 4 Maverick | 37 |
Interactive version: theaggregate.ai/benchmark?slug=carryonbench-utility-benign-clarifications · How It Works · Data refreshed daily, snapshot 2026-10-07.