Qwen4 Max vs
Claude Haiku 5.5
Both are currently SOTA. Each row is one board’s result, and the marked figure is the one that board ranks higher. Rows are not added together, so there is no overall winner: the boards measure different things, and where they differ, the difference is the finding.Each rank is the board’s own: the rank it publishes, or the model’s place in the board’s own order where it publishes none. Scores from different boards cannot be compared.
| Board | Qwen4 Max | Claude Haiku 5.5 |
|---|---|---|
| Artificial Analysis LLM Leaderboard Artificial Analysis Intelligence Index | 42.9 No. 13 of 56 | 36.5 No. 14 of 56 |
| Arena — Agent Overall | 0.075 No. 14 of 65 | -0.004 No. 36 of 65 |
| Artificial Analysis Coding Agents Artificial Analysis Coding Agent Index | — | 0.296 No. 23 of 25 |
| Epoch AI Benchmarking Hub FrontierMath Tiers 1–3 | — | 75.1% No. 24 of 88 |
| Terminal-Bench 4.0 Terminal-Bench 4.0 | — | 40.0% No. 28 of 62 |
| Arena — Code Overall | 1750 No. 6 of 157 | 1593 No. 40 of 157 |
| Arena — Text Overall (style control) | — | 1421 No. 133 of 421 |
At a glance
| Developer | Alibaba | Anthropic |
|---|---|---|
| Weights | closed | closed |
| Released | — | — |
| Price Blended Price, USD per 1M tokens (3:1 input:output) | $2.25 | $0.20 |
| Value: where it sits on the kill line | SOTA | Executioner |
Qwen4 Max →Claude Haiku 5.5 → How the kill line is drawn →