Qwen3.8 Max
Qwen3.8 Max (0902)Newer version
Gemini 4 ArgonCheaper, as good
GPT-6.1 SolBetter
MiMo v2.6 ProOpen weightsQwen3.8 MaxRankingat max effort?
Alternatives?
CheaperClaude Opus 5.5high+33%-32%$1.82No forced tool use+5BetterMiMo v2.6 ProthinkingOpen+15%-95%$0.13Pass back reasoning_content+3CheaperGemini 4 Argonhigh+31%-39%*$1.62Synchronous output stops at 65,536+2CheaperGPT-6 Astraxhigh+30%-14%$2.31Monitoring can return HTTP 403+2CheaperClaude Sonnet 5.5xhigh+29%-25%$2.01No forced tool use+5Claude Opus 5.5high+33%-32%$1.82No forced tool use+5Gemini 4 Argonhigh+31%-39%*$1.62Synchronous output stops at 65,536+2GPT-6 Astraxhigh+30%-14%$2.31Monitoring can return HTTP 403+2Claude Sonnet 5.5xhigh+29%-25%$2.01No forced tool use+5GPT-6.1 Solmax+29%-73%$0.72Remove unsupported parameters+5MiMo v2.6 ProthinkingOpen+15%-95%$0.13Pass back reasoning_content+3GPT-6.1 Terrahigh+9%-95%$0.13Claude Haiku 5.5max+8%-92%$0.21Remove sampling parameters+5GPT-6.1 Solmedium+19%-92%$0.21Remove unsupported parameters+5
MiniMax M3.1 Flashthinking+6%-92%$0.22No strict JSON+1MiMo v2.6 ProthinkingOpen+15%-95%$0.13Pass back reasoning_content+3
DeepSeek V4.1 Flash 1017maxOpen+12%-90%$0.26Pass back reasoning_content · Chat Completions+4
GLM 5.3maxOpen+12%-25%$2.01No image input+4
Step 5 PreviewthinkingOpen+9%-62%$1.03web_search not available · Chat Completions+2
Kimi K3maxOpen+9%-25%$2.00Sampling parameters are fixed+3Qwen3.8 Max (0902)thinking+13%+103%$5.41Context: 1M → 984KQwen3.8 2.4T A95BthinkingOpensame-19%$2.16Context: 1M → 262K+6Qwen3.8 FlashthinkingOpensame-86%$0.37Context: 1M → 262K+6Qwen3.7 Maxthinking-27%-65%$0.93No image inputQwen3.6 Max Previewthinking-29%—$1.30 / $7.80Context: 1M → 262K+2Qwen3.8 27BmediumOpen-31%-52%$1.27Context: 1M → 262K+5Qwen3.6 Plusthinking-33%—$0.50 / $3Qwen3.7 Plusthinking-37%-92%$0.22All Alibaba (Qwen) models →History
7 versionsBenchmarks
all categoriesWatch out 6
max_tokens range · QwenCloud›
With thinking enabled, max_tokens outside 1–32768 returns HTTP 400; use max_completion_tokens to budget reasoning plus answer without that cap. Source ↗
File is not delivered · QwenCloud / Responses›
PDF input through Responses returns HTTP 200 without delivering the file to the model; use Chat Completions or DashScope. Source ↗
Not simultaneously · QwenCloud›
Setting reasoning_effort and thinking_budget together returns an error; set only one. Source ↗
tool_choice only supports auto or none · QwenCloud›
With thinking enabled, tool_choice supports only auto or none. Source ↗
Include the word JSON · QwenCloud›
JSON Object mode requires the word JSON in a system or user message; otherwise the API returns an error. Source ↗
Agent search_strategy not supported · QwenCloud / Chat Completions›
Chat Completions does not support Qwen3.8 agent search_strategy; use the Responses web_search tool for multi-turn retrieval. Source ↗
FAQ
How much does Qwen3.8 Max cost?›
$2 per 1M input tokens and $6 per 1M output tokens; cached input $0.25 per 1M; cheapest via DeepInfra at $1.65 / $4.95.
Is Qwen3.8 Max free to use?›
There is no free API; the cheapest paid price is $1.65 per 1M input tokens.
How can I use Qwen3.8 Max?›
Through the API of Alibaba, DeepInfra, Novita.
What is the context window of Qwen3.8 Max?›
1M tokens, with up to 131K output tokens.
When was Qwen3.8 Max released? Is there a newer version?›
Released 2026-08-03. It has been superseded by Qwen3.8 Max (0902).
Is Qwen3.8 Max open source?›
No, Qwen3.8 Max is a closed model available through an API. The closest open-weights option here is MiMo v2.6 Pro.
Can Qwen3.8 Max read images?›
It accepts text and images, video as input.
What is Qwen3.8 Max?›
Qwen3.8 Max is a closed language model from Alibaba (Qwen), released Aug 2026, the successor to Qwen3.7 Max.
What are the best alternatives to Qwen3.8 Max?›
Claude Opus 5.5 (cheaper), MiMo v2.6 Pro (better).