Qwen3.8 Omni Flash overview
Qwen3.8 Omni Flash specs
- Model ID
alibaba/qwen3.8-omni-flash- Developer
- Alibaba Cloud
- Context window
- 1,000,000 tokens
- Max output
- 131,072 tokens
- Input
- Audio, Image, Text, Video
- Output
- Text
- API
- OpenAI-compatible, base URL
https://api.aimlapi.com/v1
What Is Qwen3.8 Omni Flash API?
Qwen3.8-Omni-Flash is the speed- and cost-optimised omni tier of Alibaba Cloud's Qwen3.8 family. It is natively multimodal on the way in, text, images, audio and video all go into the same prompt, and returns text. With a 1,000,000-token context window and hybrid thinking, it is aimed at high-volume multimodal understanding rather than generation.
Technical Reference
- Model ID: alibaba/qwen3.8-omni-flash
- Release date: September 17, 2026
- Context window: 1,000,000 tokens
- Max output tokens: 131,072 tokens
- Modalities: Audio, Image, Text, Video → Text
- Capabilities: hybrid thinking, parallel tool calls, structured output, streaming, vision, context caching
- Endpoint: POST /v1/chat/completions
What Qwen3.8 Omni Flash Brings to the Table
Four input modalities, one prompt
Audio, video, images and text are handled natively, so a recording, a screenshot and the surrounding instructions can be sent together without a separate transcription or captioning step.
A million tokens of context
Long enough for whole repositories, large document sets or extended transcripts, with up to 131,072 tokens back.
Hybrid thinking, switchable per request
Reasoning is off by default and turned on with enable_thinking, with thinking_budget to cap how much of it you pay for. Reasoning tokens are billed at the output rate.
Where Qwen3.8 Omni Flash Makes the Most Sense
Multimodal understanding at volume
Classifying, summarising or extracting from mixed media where a frontier model would be overkill on price.
Audio and video analysis
Meeting recordings, calls and clips go in directly and come back as structured text.
Long-document work
The 1M window removes most of the need for chunking and retrieval scaffolding.
Agentic workflows
Parallel tool calling and structured output support agent loops that plan, call tools and iterate.
Use cases
- Audio and video understanding, meetings, calls, clips
- Document and screenshot analysis at high volume
- Long-context summarisation and extraction over 1M-token inputs
- Agentic workflows with parallel tool calling and structured output
- Cost-sensitive multimodal pipelines running continuously
Capabilities
- Reasoning
- Streaming
- Structured outputs
- Function calling
- Vision
Qwen3.8 Omni Flash API pricing
| Usage | Price | Unit |
|---|---|---|
| Input | $0.15 | 1M tokens |
| Output | $0.47 | 1M tokens |
| Cached input | $0.016 | 1M tokens |
Prices from the AI/ML API price list, checked on Nov 4, 2026.
Qwen3.8 Omni Flash vs other models: token prices
AI/ML API input and output rates for 5 models, per million tokens.
| Model | Input | Output |
|---|---|---|
 Qwen3.8 Omni Flash | $0.15 | $0.47 |
| Qwen3.8-Flash | $0.16 | $0.47 |
| Qwen 3.8 27B | $0.4761 | $3.39 |
| Qwen3.8 2.4T A95B | $2.65 | $6.61 |
| Qwen3.8 Max Prime | $4.23 | $12.7 |
How to use the Qwen3.8 Omni Flash API
View docs
Build with your agent
Give your agent the model details, code and documentation.
1Get your API key
Create API keyexport AIMLAPI_KEY="8f2b4c7d9a1e6b3f…"2Make your first request
curl 'https://api.aimlapi.com/v1/chat/completions' \
-H "Authorization: Bearer $AIMLAPI_KEY" \
-H 'Content-Type: application/json' \
-d '{
"model": "alibaba/qwen3.8-omni-flash",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}'import os
import requests
headers = {
"Authorization": f"Bearer {os.environ['AIMLAPI_KEY']}",
"Content-Type": "application/json",
}
payload = {
"model": "alibaba/qwen3.8-omni-flash",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}
response = requests.post("https://api.aimlapi.com/v1/chat/completions", headers=headers, json=payload, timeout=120)
response.raise_for_status()
print(response.json())const response = await fetch("https://api.aimlapi.com/v1/chat/completions", {
method: "POST",
headers: {
Authorization: `Bearer ${process.env.AIMLAPI_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "alibaba/qwen3.8-omni-flash",
"messages": [
{
"role": "user",
"content": "Write a haiku about the ocean."
}
]
}),
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
console.log(await response.json());Models similar to Qwen3.8 Omni Flash
Qwen3.8 Omni Flash FAQ
What is the context window of Qwen3.8 Omni Flash?
Qwen3.8 Omni Flash has a 1,000,000 token context window and can return up to 131,072 tokens.
What inputs and outputs does Qwen3.8 Omni Flash support?
It accepts audio, images, text and video as input and returns text. All four input types can go into the same prompt, so a recording or a clip does not need a separate transcription step first.
How much does Qwen3.8 Omni Flash cost?
Qwen3.8 Omni Flash costs $0.15 input and $0.47 output per 1M tokens on AI/ML API, billed as you go with no subscription.
Does Qwen3.8 Omni Flash generate audio?
No. Audio and video are input modalities only, the model listens and watches, but answers in text. There is no parameter on the chat completions endpoint to request an audio response.
What is hybrid thinking and how do I turn it on?
Reasoning is off by default. Set enable_thinking to true to switch it on, and use thinking_budget to cap how many reasoning tokens a request may spend. Those tokens are billed at the output rate.
Does Qwen3.8 Omni Flash support tool calling and structured output?
Yes. It supports function calling including parallel tool calls, structured (JSON schema) output, streaming and context caching.
Who created Qwen3.8 Omni Flash?
Qwen3.8 Omni Flash was built by Alibaba Cloud as the speed- and cost-optimised omni tier of the Qwen3.8 family. It became available on September 17, 2026.
How do I call Qwen3.8 Omni Flash through AI/ML API?
It is served on the OpenAI-compatible chat completions endpoint. Point any OpenAI SDK at https://api.aimlapi.com/v1 and set the model to alibaba/qwen3.8-omni-flash. The short alias qwen3.8-omni-flash also resolves to it.
What is the Qwen3.8 Omni Flash model ID?
Use alibaba/qwen3.8-omni-flash as the model parameter. Requests go to the OpenAI-compatible endpoint with your AI/ML API key.