| Metric | Value |
|---|---|
| Arize AI wins | 105 |
| Braintrust wins | 720 |
| Abstains (no tool) | 90 |
| Other tool chosen | 1529 |
| Decisive cases | 825 |
| Arize AI win rate (unweighted) | 12.7% |
| 95% CI | 10.6% - 15.2% |
| Arize AI win rate (weighted) | 12.7% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Arize AI | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| GPT 5.3 Codex | Frontier | 0 | 126 | 0 | 6 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 113 | 0 | 19 | 0% |
| Kimi K2.5 | Frontier | 0 | 109 | 3 | 7 | 0% |
| Claude Haiku 4.5 | Small | 0 | 103 | 1 | 21 | 0% |
| GPT 5.4 | Frontier | 0 | 92 | 0 | 40 | 0% |
| GLM 5 Turbo | Frontier | 0 | 84 | 19 | 29 | 0% |
| Claude Sonnet 4.6 | Frontier | 0 | 58 | 0 | 74 | 0% |
| Llama 4 Scout | Small | 40 | 0 | 4 | 77 | 100% |
| MiniMax M2.7 | Frontier | 1 | 31 | 5 | 92 | 3% |
| Gemini 2.5 Flash | Small | 27 | 0 | 1 | 99 | 100% |
| Llama 4 Maverick | Frontier | 13 | 0 | 0 | 114 | 100% |
| Gemini 2.5 Pro | Frontier | 10 | 0 | 9 | 113 | 100% |
| Devstral 2 2512 | Mid | 8 | 0 | 4 | 113 | 100% |
| DeepSeek R1 0528 | Frontier | 3 | 0 | 7 | 122 | 100% |
| GPT 5.4 Mini | Mid | 0 | 3 | 3 | 126 | 0% |
| Mistral Small 4 | Mid | 2 | 0 | 1 | 121 | 100% |
| Qwen3 Coder Next | Mid | 1 | 0 | 3 | 127 | 100% |
| MiMo V2 Pro | Frontier | 0 | 1 | 8 | 123 | 0% |
| DeepSeek V3.2 | Mid | 0 | 0 | 22 | 106 | n/a |
| Prompt | Tier | Arize AI | Braintrust | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Advanced | 23 | 153 | 2 | 222 | 13% |
| ai-revenue-ops-copilot | Beginner | 15 | 143 | 10 | 242 | 9% |
| ai-support-agent-platform | Advanced | 18 | 123 | 5 | 264 | 13% |
| ai-revenue-ops-copilot | Intermediate | 11 | 122 | 4 | 267 | 8% |
| ai-support-agent-platform | Beginner | 18 | 92 | 64 | 237 | 16% |
| ai-support-agent-platform | Intermediate | 20 | 87 | 5 | 297 | 19% |
