Cresta-Bench: leaderboard
A model runs four contact-center agents taking calls from an LLM-simulated caller across 181 cases, each run three times at the model's lowest reasoning setting, except Claude Sonnet 5.5 at low; a case counts only when all three runs pass, from Cresta's own runs. It rewards getting the same case right every time, not once.
Metric: Cases passed on all 3 runs, mean over 4 agents (%). Source: fireworks.ai. Saturation forecast: Around February 2027. 14 models tracked. Updated 16 Nov 2026.
Top models
| # | Model | Setting | Score | Overall rank |
|---|---|---|---|---|
| 1 | Claude Opus 5.5 | Low | 82.57 | #5 |
| 2 | Claude Sonnet 5.5 | Low | 78.04 | #19 |
| 3 | GPT-6 Sol | Non-reasoning | 77.94 | #26 |
| 4 | Gemini 3.8 Flash | Low | 76.14 | #36 |
| 5 | GPT-5.4 | Non-reasoning | 65.14 | #128 |
| 6 | Gemini 3.6 Flash | Minimal | 50.19 | #75 |
| 7 | GPT-5.2 | Non-reasoning | 49.47 | #178 |
| 8 | Qwen 3.8 Max | Non-reasoning | 48.67 | #59 |
| 9 | GLM-5.3 | Low | 43.06 | #68 |
| 10 | GPT-5.6 Luna | Non-reasoning | 38.81 | #122 |
| 11 | Kimi K3 | Low | 34.85 | #56 |
| 12 | GPT-6 Luna | Non-reasoning | 32.73 | #99 |
| 13 | DeepSeek V4.1 Flash (1017) | Non-reasoning | 32.64 | #41 |
| 14 | Claude Haiku 5.5 | Non-reasoning | 20.05 | #70 |
No result here: #1 Claude Fable 5.2, #2 Gemini 4 Carbon, #6 GPT-6 Astra.
Interactive version: theaggregate.ai/benchmark?slug=cresta-bench · How It Works · Data refreshed daily, snapshot 2026-11-16.