Leading: Claude Code (55.3%)
| Metric | Value |
|---|---|
| Intercom wins | 117 |
| Claude Code wins | 145 |
| Abstains (no tool) | 103 |
| Other tool chosen | 3153 |
| Decisive cases | 262 |
| Intercom win rate (unweighted) | 44.7% |
| 95% CI | 38.8% - 50.7% |
| Intercom win rate (weighted) | 44.7% |
Verified critics can leave comments here.
Verified critics can leave comments here.
| Model | Tier | Intercom | Claude Code | None | Other | A rate |
|---|---|---|---|---|---|---|
| MiniMax M2.7 | Frontier | 7 | 42 | 0 | 65 | 14% |
| Claude Haiku 4.5 | Small | 7 | 34 | 0 | 147 | 17% |
| DeepSeek V3.2 | Mid | 12 | 14 | 0 | 99 | 46% |
| Llama 4 Scout | Small | 23 | 0 | 7 | 149 | 100% |
| Mistral Small 4 | Mid | 19 | 0 | 14 | 119 | 100% |
| MiniMax M3 | Frontier | 0 | 19 | 0 | 43 | 0% |
| Claude Sonnet 4.6 | Frontier | 16 | 0 | 0 | 173 | 100% |
| GLM 5 Turbo | Frontier | 0 | 14 | 2 | 109 | 0% |
| GPT 5.4 Mini | Mid | 9 | 0 | 1 | 177 | 100% |
| GPT 5.4 | Frontier | 8 | 0 | 0 | 118 | 100% |
| MiMo V2.5 Pro | Frontier | 2 | 5 | 4 | 51 | 29% |
| Kimi K2.7 Code | Frontier | 0 | 7 | 2 | 52 | 0% |
| Llama 4 Maverick | Frontier | 5 | 0 | 21 | 158 | 100% |
| DeepSeek V4 Flash | Mid | 3 | 2 | 8 | 46 | 60% |
| GPT 5.5 | Frontier | 0 | 3 | 0 | 54 | 0% |
| Gemini 2.5 Pro | Frontier | 2 | 0 | 1 | 185 | 100% |
| MiMo V2 Pro | Frontier | 2 | 0 | 0 | 122 | 100% |
| GPT 5.3 Codex | Frontier | 0 | 2 | 0 | 186 | 0% |
| DeepSeek R1 0528 | Frontier | 1 | 0 | 18 | 167 | 100% |
| Kimi K2.5 | Frontier | 1 | 0 | 1 | 111 | 100% |
| Claude Opus 4.8 | Frontier | 0 | 1 | 0 | 61 | 0% |
| DeepSeek V4 Pro | Frontier | 0 | 1 | 9 | 53 | 0% |
| Gemini 3.5 Flash | Small | 0 | 1 | 1 | 59 | 0% |
| Claude Opus 4.6 | Frontier | 0 | 0 | 0 | 126 | n/a |
| Devstral 2 2512 | Mid | 0 | 0 | 12 | 150 | n/a |
| Gemini 2.5 Flash | Small | 0 | 0 | 0 | 126 | n/a |
| GLM 5.2 | Frontier | 0 | 0 | 1 | 62 | n/a |
| Qwen3 Coder Next | Mid | 0 | 0 | 1 | 185 | n/a |
| Prompt | Tier | Intercom | Claude Code | None | Other | A rate |
|---|---|---|---|---|---|---|
| ai-support-agent-platform | Beginner | 116 | 5 | 1 | 386 | 96% |
| ai-revenue-ops-copilot | Beginner | 0 | 37 | 0 | 465 | 0% |
| ai-revenue-ops-copilot | Advanced | 0 | 34 | 3 | 477 | 0% |
| ai-support-agent-platform | Advanced | 0 | 20 | 1 | 508 | 0% |
| ai-revenue-ops-copilot | Intermediate | 0 | 17 | 1 | 511 | 0% |
| ai-agent-application | Intermediate | 0 | 15 | 36 | 81 | 0% |
| ai-support-agent-platform | Intermediate | 1 | 7 | 0 | 523 | 13% |
| ai-agent-application | Advanced | 0 | 7 | 22 | 106 | 0% |
| ai-agent-application | Beginner | 0 | 3 | 39 | 96 | 0% |
