Qwen 3.8 Max
Qwen 3.8 Max is an API-only (closed weights) language model from Qwen, released July 19, 2026. Its makers disclosed about 2.4T parameters. The context window is 1M tokens. On the Epoch Capabilities Index it scores 157.6, rank 29 of 256 rated models. API price: $2 input / $6 output per million tokens.
Today, we are officially releasing Qwen 3.8-Max, the most capable model in the Qwen family to date. This also marks the first time we will open-source the weights of a Qwen-Max-class model — the open weights will be released next week.
Check it against your own PC →
Strengths by category
ModelTape's 0–100 fit on Epoch AI benchmark results.
| Category | Score | Based on |
|---|---|---|
| Reasoning | 24.9 | Chess Puzzles, Mystery Game Puzzles |
| Math | 60.4 | ProofBench, OTIS Mock AIME 2024-2025, FrontierMath-Tiers-1-3-v2-Private, FrontierMath-Tier-4-v2-Private |
| Coding | 57.4 | MirrorCode, FrontierSWE, DeepSWE |
| Knowledge | 75.6 | GPQA diamond, SimpleQA Verified |
Benchmark results
| Benchmark (Epoch AI) | Score |
|---|---|
| OTIS Mock AIME | 99.4% |
| GPQA Diamond | 94.0% |
| FrontierMath (Tiers 1–3) | 78.7% |
| ProofBench | 58.0% |
| DeepSWE | 57.5% |
| MirrorCode | 52.2% |
| FrontierMath-Tier-4-v2-Private | 46.3% |
| SimpleQA Verified | 45.8% |
| Chess Puzzles | 43.2% |
| Mystery Game Puzzles | 31.7% |
| FrontierSWE | 15.8% |
Links
Announcement · Epoch AI benchmarks
Models with a similar score
Kimi K3.1 Pro · GPT-5.6 Terra · Claude Haiku 5.5 · GPT-5.4 Pro · GPT-5.5 · Claude Opus 4.8 · DeepSeek V4.1 Flash 1017 · Gemini 3.8 Flash
More from Qwen
Qwen3.8 Omni Flash · Qwen3.8 Max (0902) · Qwen3.8 Flash · Qwen3.8 Flash Next · Qwen 3.8 27B · Qwen3.8-2.4T-A95B · Qwen3.7 Flash · Qwen3.7-Plus