OpenCompass LLM - Agent (CompassBench 2404): leaderboard
OpenCompass CompassBench 2404 edition (file dated 2024-05-09): agent (tool use) category score. CompassBench is OpenCompass's periodic evaluation of chat models on a private question set; each edition is a separate board. Edition 2405 republished 20 of these results for the same task set; those models appear on that board only. Score 0-100, higher is better.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Estimated already saturated. 31 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 Turbo (Preview) | 82 | #416 |
| 2 | Qwen-72B-Chat | 58.1 | #835 |
| 3 | Qwen-14B-Chat | 54.9 | #1161 |
| 4 | Yi Chat 34B | 53.6 | #950 |
| 5 | Qwen-7B Chat | 48.5 | #1353 |
| 6 | vicuna-13B-v1.5-16k | 43.2 | #1417 |
| 7 | OrionStar-Yi-34B-Chat | 42.3 | #872 |
| 8 | Yi 6B (Chat) | 40.9 | #1379 |
| 9 | deepseek-llm-7B-chat | 40.5 | #1320 |
| 10 | Llama 2 70B Chat | 38.6 | #1275 |
| 11 | Vicuna-7B-v1.5-16k | 37.9 | #1470 |
| 12 | Baichuan2-7B-Chat | 37.7 | #1384 |
| 13 | Mistral 7B Instruct (v0.2) | 37.3 | #1311 |
| 14 | WizardLM-70B-V1.0 | 35.6 | #1219 |
| 15 | WizardLM-13B-V1.2 | 34.9 | #1399 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=opencompass-llm-agent-compassbench-2404 · How It Works · Data refreshed daily, snapshot 2026-11-05.