Qwen3.8 Flash
Qwen3.8 FlashRankingin thinking mode?
Alternatives?
CheaperGPT-6.1 Solhigh+35%-14%$0.32Remove unsupported parameters+5
BetterClaude Haiku 5.5xhigh+12%-67%$0.12Remove sampling parameters+5
OpenGLM 5.3 Flashmax+5%-24%$0.28Memory (4-bit): 108 GB → 193 GB+5
CheaperMiMo v2.6 ProthinkingOpen+30%-64%$0.13Memory (4-bit): 108 GB → 615 GB+4
CheaperMiniMax M3.1 Flashthinking+26%-41%$0.22Complete model response must be appendedGPT-6.1 Solhigh+35%-14%$0.32Remove unsupported parameters+5MiMo v2.6 ProthinkingOpen+30%-64%$0.13Memory (4-bit): 108 GB → 615 GB+4MiniMax M3.1 Flashthinking+26%-41%$0.22Complete model response must be appendedClaude Haiku 5.5max+17%-43%$0.21Remove sampling parameters+5
DeepSeek V4.1 Flash 1017maxOpen+14%-30%$0.26Memory (4-bit): 108 GB → 458 GB+5Claude Haiku 5.5xhigh+12%-67%$0.12Remove sampling parameters+5GPT-6.1 Sollow+24%-65%$0.13Remove unsupported parameters+5MiMo v2.6 ProthinkingOpen+30%-64%$0.13Memory (4-bit): 108 GB → 615 GB+4MiniMax M3.1 Flashthinking+26%-41%$0.22Complete model response must be appendedGPT-6 Solmedium+8%-34%$0.25GLM 5.3 FlashmaxOpen+5%-24%$0.28Memory (4-bit): 108 GB → 193 GB+5MiniMax M3thinkingOpensame+36%$0.51Memory (4-bit): 108 GB → 256 GB+2MMotif 3Open-3%+280%$0.50 / $2Memory (4-bit): 108 GB → 189 GB+1MiMo v2.6 FlashthinkingOpen-8%-83%$0.062Memory (4-bit): 108 GB → 186 GB+4Qwen3.8 27BxhighOpen-11%+171%$1.01max_tokens range · QwenCloud+4Qwen3.8 Max (0902)thinking+13%+1,353%$5.41Qwen3.8 2.4T A95BthinkingOpen+12%+479%$2.16Memory (4-bit): 108 GB → 1,468 GB+6Qwen3.7 Maxthinking+6%+208%$1.15No image inputQwen3.7 Plusthinking-6%-42%$0.22Qwen3.6 Max Previewthinking-19%—$1.30 / $7.80No image input+1Qwen3.5 397B A17BthinkingOpen-24%+28%$0.47Memory (4-bit): 108 GB → 242 GBQwen3 Max Thinkingthinking-26%—$1.20 / $6Context: 262K → 256K+1Qwen3.6 Plusthinking-27%—$0.50 / $3All Alibaba (Qwen) models →Run it yourself
estimateBenchmarks
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 ↗
Smaller values automatically adjusted · QwenCloud›
In Flash thinking mode, temperature below 0.6 is automatically raised to 0.6. Source ↗
Static YaRN affects shorter texts · Self-hosted›
Static YaRN can hurt short-text performance; change rope_parameters only when longer contexts are required. 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 ↗
FAQ
How much does Qwen3.8 Flash cost?›
$0.15 per 1M input tokens and $0.47 per 1M output tokens; cheapest via DeepInfra at $0.11 / $0.38. A typical benchmark task costs about $0.37 at default effort.
Is Qwen3.8 Flash free to use?›
There is no free API; the cheapest paid price is $0.11 per 1M input tokens. The weights are open, so you can run it yourself.
How good is Qwen3.8 Flash, and what is it best at?›
#52 of 334 in Chat & Knowledge, #26 of 132 in Coding, #30 of 133 in Agents & Tool Use, #21 of 124 in Vision & Documents. Strongest at: Coding, Agents & work tasks, Reasoning, Science, Long context, Vision.
Is Qwen3.8 Flash better than Grok Build 0.1?›
On Humanity's Last Exam, Qwen3.8 Flash scores 38 vs 38.3 — about the same.
How can I use Qwen3.8 Flash?›
Through the API of Alibaba, DeepInfra, Featherless, Novita.
What is the context window of Qwen3.8 Flash?›
262K tokens, with up to 131K output tokens.
When was Qwen3.8 Flash released? Is there a newer version?›
Released 2026-08-26. It is the current version in its line.
Is Qwen3.8 Flash open source?›
Yes, the weights are open (qwen-community-1.0 licence).
Can I run Qwen3.8 Flash locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
Can Qwen3.8 Flash read images?›
It accepts text and images, video as input.
What is Qwen3.8 Flash?›
Qwen3.8 Flash is an open-weights language model from Alibaba (Qwen), released Aug 2026.
What are the best alternatives to Qwen3.8 Flash?›
GPT-6.1 Sol (cheaper), Claude Haiku 5.5 (better), GLM 5.3 Flash (open weights).