| Metric | Value |
|---|---|
| Metabase wins | 609 |
| Mixpanel wins | 1522 |
| Abstains (no tool) | 459 |
| Other tool chosen | 5375 |
| Decisive cases | 2131 |
| Metabase win rate (unweighted) | 28.6% |
| 95% CI | 26.7% - 30.5% |
| Metabase win rate (weighted) | 28.6% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Metabase | Mixpanel | None | Other | A rate |
|---|---|---|---|---|---|---|
| Llama 4 Maverick | Frontier | 0 | 284 | 0 | 135 | 0% |
| Devstral 2 2512 | Mid | 1 | 200 | 0 | 202 | 0% |
| Claude Sonnet 4.6 | Frontier | 84 | 86 | 1 | 249 | 49% |
| Gemini 2.5 Flash | Small | 0 | 167 | 22 | 138 | 0% |
| Claude Haiku 4.5 | Small | 12 | 143 | 5 | 246 | 8% |
| MiMo V2 Pro | Frontier | 57 | 84 | 25 | 164 | 40% |
| Qwen3 Coder Next | Mid | 86 | 41 | 5 | 288 | 68% |
| DeepSeek R1 0528 | Frontier | 19 | 93 | 149 | 157 | 17% |
| DeepSeek V3.2 | Mid | 3 | 109 | 27 | 188 | 3% |
| Claude Opus 4.6 | Frontier | 52 | 49 | 1 | 228 | 51% |
| Mistral Small 4 | Mid | 18 | 55 | 3 | 327 | 25% |
| Llama 4 Scout | Small | 0 | 69 | 25 | 324 | 0% |
| GPT 5.3 Codex | Frontier | 45 | 23 | 1 | 350 | 66% |
| MiniMax M2.7 | Frontier | 37 | 29 | 36 | 217 | 56% |
| GPT 5.4 Mini | Mid | 30 | 16 | 7 | 366 | 65% |
| GPT 5.4 | Frontier | 16 | 19 | 0 | 292 | 46% |
| MiMo V2.5 Pro | Frontier | 17 | 17 | 11 | 45 | 50% |
| Kimi K2.5 | Frontier | 30 | 3 | 23 | 255 | 91% |
| GLM 5 Turbo | Frontier | 28 | 4 | 65 | 233 | 88% |
| Gemini 2.5 Pro | Frontier | 10 | 14 | 26 | 366 | 42% |
| DeepSeek V4 Flash | Mid | 13 | 7 | 4 | 64 | 65% |
| Kimi K2.7 Code | Frontier | 14 | 4 | 3 | 69 | 78% |
| GPT 5.5 | Frontier | 13 | 2 | 1 | 71 | 87% |
| DeepSeek V4 Pro | Frontier | 12 | 2 | 15 | 60 | 86% |
| GLM 5.2 | Frontier | 7 | 0 | 1 | 82 | 100% |
| Claude Opus 4.8 | Frontier | 5 | 0 | 0 | 84 | 100% |
| Gemini 3.5 Flash | Small | 0 | 1 | 0 | 89 | 0% |
| MiniMax M3 | Frontier | 0 | 1 | 3 | 86 | 0% |
| Prompt | Tier | Metabase | Mixpanel | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-revenue-ops-copilot | Beginner | 261 | 63 | 5 | 191 | 81% |
| multi-tenant-crm | Intermediate | 138 | 128 | 34 | 230 | 52% |
| fitness-tracking-app | Intermediate | 0 | 221 | 16 | 300 | 0% |
| multi-tenant-crm | Beginner | 23 | 151 | 47 | 313 | 13% |
| ai-revenue-ops-copilot | Advanced | 82 | 85 | 16 | 323 | 49% |
| ai-revenue-ops-copilot | Intermediate | 9 | 158 | 3 | 359 | 5% |
| saas-application | Beginner | 0 | 140 | 16 | 381 | 0% |
| fitness-tracking-app | Beginner | 0 | 137 | 1 | 398 | 0% |
| multi-tenant-crm | Advanced | 90 | 25 | 37 | 384 | 78% |
| saas-application | Intermediate | 0 | 105 | 52 | 378 | 0% |
| fitness-tracking-app | Advanced | 5 | 81 | 13 | 427 | 6% |
| saas-application | Advanced | 1 | 80 | 30 | 425 | 1% |
| url-shortener | Intermediate | 0 | 62 | 75 | 398 | 0% |
| url-shortener | Beginner | 0 | 45 | 109 | 382 | 0% |
| url-shortener | Advanced | 0 | 41 | 5 | 486 | 0% |