Qwen3.8 Omni Flash

Name: Qwen3.8 Omni Flash model facts
Creator: OSMEXO

alibaba/qwen3.8-omni-flash

Alibaba
Family · Qwen3 Omni Flash
Text · Image · Audio · Video → Text
Tools · JSON · Vision

Not on OSMEXO yet

Request this model

Spec & list price

Context1M

Max output128K

Input / 1M$0.15

Output / 1M$0.473.1× input

Usage

Tokens / day4.3B▲ 44% 7d

Usage rank#121

Requests / day504.2K

Tool calls / req0.10

Per request · 30 days

Avg input12.0K0.8× median

Avg output1.7K2.4× median

Cache hit39%0.6× median

Reasoning share76%1.4× median

Tool-call failures12%6.4× median

Consumption

Peak 11.9B 30 OctLow 2.0B 27 SepRank today #121

News & trends

Trending · 14 days

Ep. 108: Alibaba ships Qwen3.8-Omni-Flash to watch, listen and call tools (Daily Recap, Sep 18)3 views

News · latest from Alibaba

Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hardgithub.com · 31 Oct
Qwen3.8-Flash Now Available in the EU on Alibaba Cloudtwitter.com · 31 Oct
Alibaba's Qwen3.8-Flash-Next model tried to upload a file to Beijing in a testnotebookcheck.net · 27 Oct
Neuron Explorer for Qwen3.5-4B: See every neuron at work, make targeted editsgithub.com · 26 Oct

Speed

Output speed#160/33464 tok/sabout median

First token#2893.28 s3.1× longer than median

Technical details

Tokenizer: Qwen
Moderation: none

Request parameters 16

tools
tool_choice
response_format
structured_outputs
reasoning
include_reasoning
temperature
top_p
top_k
max_tokens

+6 more

stop
seed
frequency_penalty
logprobs
presence_penalty
top_logprobs

When to use

Best for

Audio and video understanding in apps
Mixed media input in one request
Image understanding in chat apps
Fast low-cost multimodal assistants

Think twice for

Speech or audio output (text-only output)
Image generation
Hardest reasoning tasks
Self-hosting (closed weights)

Same family

Qwen3 Omni Flash

Qwen3.5 Omni Flash
earlier · Mar 2026
12.5

Want Qwen3.8 Omni Flash on OSMEXO?

Request this model on OSMEXO