Qwen3.8 Max
Alibaba's proprietary top-end model and the ceiling of the Qwen line, reserved for the hardest reasoning and agentic tasks. Unlike most of the family, its weights are not released.
At a glance
Released
Aug 12, 2026
Provider
Alibaba
Context window
256,000 tokens
Max output
32,000 tokens
Modalities
Text + image → text
Knowledge cutoff
May 2026
API model ID
qwen3.8-maxLicense
Proprietary (API only)
Pricing
Standard
Input$2 / 1M
Output$6 / 1M
* Rates are from a third-party aggregator — verify before relying on them. A typical 1,500-token prompt with a 650-token answer costs ≈ $0.007 on the standard tier. Run your own prompt through the calculator →
What's new vs the previous model
Compared withQwen3.7 Max (April 2026)
- Stronger reasoning and coding than Qwen3.7 Max
- Better multilingual and long-context performance
- Remains API-only, unlike the open Plus and Flash tiers
Best use cases
Complex reasoning
Tackles demanding analytical and math-heavy problems.
Agent orchestration
Coordinates multi-tool agents with reliable planning.
Multilingual work
Delivers strong quality across a wide range of languages.
Strengths & limitations
Strengths
- Top-tier reasoning
- Strong multilingual
- Large context
Limitations
- Closed weights, no self-hosting
- Most expensive in the Qwen line
How to test it
- No-code: open chat.qwen.ai and pick Qwen3.8 Max from the model switcher — no API key needed.
- Playground: sign in to Alibaba Cloud Model Studio and run prompts in the workbench, then copy the generated code.
- API: create a key in Alibaba Cloud Model Studio and call the API (example below).
curl https://dashscope-intl.aliyuncs.com/compatible-mode/v1/chat/completions \
# OpenAI-compatible endpoint
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "qwen3.8-max",
"messages": [{"role": "user", "content": "Explain tokenization in one paragraph."}]
}'Where to get access
Developer docs
API reference, SDKs and quickstarts for Alibaba.
www.alibabacloud.com/help/en/model-studio ↗