qwen3.8-maxQwen3.8 Max is Alibaba’s newest Qwen flagship. It is reasoning-first — thinking is on by default and its depth is graded through reasoning_effort (low / medium / xhigh), so you trade latency and output cost against answer quality per request. 1M-token context, tool calling, JSON output and streaming, served OpenAI-compatible via /v1/chat/completions. We do not add a content-moderation layer of our own on top of this model: prompts and completions pass through as sent. The model's own safety systems still apply upstream — a request can come back refused as data_inspection_failed when the input is judged inappropriate — so this is not an unfiltered endpoint. Reviewing what you generate, and complying with the law and your own platform's rules, is your responsibility.
View complete API reference with all parameters and examples.
View complete API reference with streaming, thinking, and more.
Billing: Cost = (input_tokens * input_price + output_tokens * output_price) / 1,000,000
reasoning_effort low / medium / xhigh trades depth against latency and output cost
Million-token context window
Reasoning tokens are part of completion_tokens — typically 85-95% of the output on short prompts. Set reasoning_effort to control it.
Function calling + JSON output, OpenAI-compatible
Qwen3.8 Max is a Large Language Model API provided by Alibaba. Qwen3.8 Max is Alibaba’s newest Qwen flagship. It is reasoning-first — thinking is on by default and its depth is graded through reasoning_effort (low / medium / xhigh), so you trade latency and output cost against answer quality per request. 1M-token context, tool calling, JSON output and streaming, served OpenAI-compatible via /v1/chat/completions. We do not add a content-moderation layer of our own on top of this model: prompts and completions pass through as sent. The model's own safety systems still apply upstream — a request can come back refused as data_inspection_failed when the input is judged inappropriate — so this is not an unfiltered endpoint. Reviewing what you generate, and complying with the law and your own platform's rules, is your responsibility. Through APIMODELS platform, you can access this model via a unified API with transparent pay-as-you-go pricing. Current pricing: Input: $2.00, Output: $6.00 per 1M tokens.
Build intelligent conversational systems to automatically answer user queries and improve service efficiency.
Automatically write articles, emails, ad copy, and other text content to boost productivity.
Assist with code writing, debugging, and code review to accelerate software development.
Understand and analyze unstructured data, extract key insights, and generate summary reports.
Qwen3.8 Max is available through APIMODELS at: Input: $2.00, Output: $6.00 per 1M tokens. Billing is pay-as-you-go — you only pay for what you generate.
Sign up at APIMODELS, get your API key, and call our unified API endpoint. We provide detailed API documentation with code examples in cURL, Python, and Node.js.
APIMODELS offers the same Qwen3.8 Max model through our aggregation platform. We provide a unified API interface so you do not need separate accounts for each provider - one API key to access all models.
Qwen3.8 Max is Alibaba's newest flagship. The main change over 3.7 Max is that thinking depth is now a graded dial: pass reasoning_effort as low, medium or xhigh and trade answer depth against latency and output cost per request, where 3.7 could only switch thinking on and off or cap its tokens. On price the newer model is the more affordable one — 3.8 Max is $2.00 / $6.00 against 3.7 Max at $2.25 / $6.75. Same endpoint; switching is a one-field change.
No — it is the same as official list price: $2.00 per 1M input, $6.00 per 1M output, cached input $0.25. We do not play a pricing game on this generation. The reasons to use us here are different ones: no Alibaba Cloud account or identity verification, reachable directly from mainland China, one key that also reaches Claude, Gemini, GPT, Grok and everything else on a single balance, pay-as-you-go with no minimum top-up, and failed requests never billed. If the lowest possible price is all you want, going direct costs the same.
Valid values are low, medium and xhigh, and the default is xhigh — the deepest. For cross-vendor compatibility, high and max map to xhigh, minimal maps to low, and none turns thinking off. It cannot be combined with thinking_budget; sending both is an error, so pick one. When reasoning_effort is not set, thinking_budget maps back to a tier: 0–4096 is low, 4097–16384 is medium, 16385–262144 is xhigh.
Because preserve_thinking is on by default for qwen3.8-max: previous turns’ reasoning_content is replayed back to the model as context. The upside is that the model remembers how it reasoned; the cost is that that historical thinking is re-billed as INPUT on every follow-up turn. Input counts on long chats therefore run well above what the visible messages suggest. Trim the history, or use a tier that does not replay thinking by default.
POST https://api.apimodels.app/v1/chat/completions with Authorization: Bearer YOUR_API_KEY and body {"model":"qwen3.8-max","messages":[{"role":"user","content":"..."}],"reasoning_effort":"low"}. Plain OpenAI shape — change base_url on openai-python or openai-node and it works. 1M-token context; add stream:true for streaming.
On APIMODELS, Qwen3.8 Max runs alongside 60+ models on one API key and one balance, so choosing is about fit, not lock-in. It supports Reasoning / Thinking, Flagship, 1M Context, Tool Calling, Streaming, and you can weigh it on price and capability against other Large Language Model models, then switch by changing a single model-name string — no new account or integration. Browse every Large Language Model option with live pricing at apimodels.app/models.
Qwen3.8 Max supports: Reasoning / Thinking, Flagship, 1M Context, Tool Calling, Streaming. See the APIMODELS docs for full parameters and call examples.
Yes. APIMODELS exposes Qwen3.8 Max through a single unified API and one key — no separate provider accounts, and no need to handle each provider's regional network access yourself.
We support Stripe (Visa, Mastercard, and other international cards) and Alipay. Credits are available instantly after payment.