CCR-Bench - Workflow Task Success: leaderboard
Metric: Task success rate (0-1, times 100): share of tasks executed exactly as specified with every objective met, on CCR-Bench's 70 logical workflow control tasks (multi-turn tool use with conditional branching, implicit nested workflows, implicit tool calls and long tool chains, checked by verification scripts), temperature 0; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro | 70 | #145 |
| 2 | GPT-4.1 | 52.9 | #240 |
| 3 | O3 Mini | 51.4 | #266 |
| 4 | DeepSeek R1 0528 | 40 | #217 |
| 5 | QwQ-32B | 38.6 | #410 |
| 6 | Qwen 3 32B (Thinking) | 38.6 | #424 (Qwen 3 32B) |
| 7 | Qwen 2.5 72B Instruct | 32.9 | #436 |
| 8 | DeepSeek V3 (0324) | 30 | #332 |
| 9 | Qwen 3 32B (Non-reasoning) | 15.7 | #424 (Qwen 3 32B) |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=ccr-bench-workflow-task-success · How It Works · Data refreshed daily, snapshot 2026-10-11.