| Metric | Value |
|---|---|
| Pusher wins | 497 |
| Supabase wins | 630 |
| Abstains (no tool) | 40 |
| Other tool chosen | 2230 |
| Decisive cases | 1127 |
| Pusher win rate (unweighted) | 44.1% |
| 95% CI | 41.2% - 47.0% |
| Pusher win rate (weighted) | 44.1% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Pusher | Supabase | None | Other | A rate |
|---|---|---|---|---|---|---|
| DeepSeek V3.2 | Mid | 9 | 140 | 0 | 31 | 6% |
| MiMo V2 Pro | Frontier | 76 | 21 | 0 | 82 | 78% |
| Claude Sonnet 4.6 | Frontier | 7 | 83 | 0 | 90 | 8% |
| MiniMax M2.7 | Frontier | 10 | 78 | 0 | 90 | 11% |
| Qwen3 Coder Next | Mid | 74 | 8 | 0 | 97 | 90% |
| Gemini 2.5 Pro | Frontier | 79 | 1 | 0 | 100 | 99% |
| GLM 5 Turbo | Frontier | 68 | 9 | 3 | 100 | 88% |
| Mistral Small 4 | Mid | 30 | 40 | 0 | 109 | 43% |
| GPT 5.4 Mini | Mid | 5 | 54 | 2 | 115 | 8% |
| Claude Opus 4.6 | Frontier | 27 | 26 | 0 | 126 | 51% |
| GPT 5.3 Codex | Frontier | 2 | 48 | 0 | 130 | 4% |
| GPT 5.4 | Frontier | 14 | 35 | 0 | 128 | 29% |
| DeepSeek R1 0528 | Frontier | 30 | 14 | 30 | 105 | 68% |
| Llama 4 Scout | Small | 31 | 2 | 0 | 145 | 94% |
| Claude Haiku 4.5 | Small | 1 | 31 | 0 | 143 | 3% |
| Gemini 2.5 Flash | Small | 7 | 23 | 0 | 150 | 23% |
| Llama 4 Maverick | Frontier | 23 | 0 | 0 | 157 | 100% |
| Kimi K2.5 | Frontier | 2 | 17 | 5 | 154 | 11% |
| Devstral 2 2512 | Mid | 2 | 0 | 0 | 178 | 100% |
| Prompt | Tier | Pusher | Supabase | None | Other | A rate |
|---|---|---|---|---|---|---|
| social-media-platform | Beginner | 91 | 97 | 17 | 168 | 48% |
| project-management-tool | Beginner | 53 | 127 | 4 | 195 | 29% |
| social-media-platform | Intermediate | 115 | 62 | 4 | 196 | 65% |
| project-management-tool | Intermediate | 47 | 119 | 3 | 208 | 28% |
| social-media-platform | Advanced | 99 | 26 | 3 | 251 | 79% |
| project-management-tool | Advanced | 27 | 69 | 0 | 283 | 28% |
| chat-application | Intermediate | 12 | 81 | 5 | 280 | 13% |
| chat-application | Advanced | 41 | 17 | 1 | 317 | 71% |
| chat-application | Beginner | 12 | 32 | 3 | 332 | 27% |