Qwen Flagship multimodal reasoning
Qwen 3.8 Max
Give complex work the context, tools and output contract it deserves.
Qwen 3.8 Max is Qwen's flagship multimodal reasoning route in the TokenHot catalog. It accepts text, images and video, with a 1M context value and up to 128K output listed in the model description. Function calling, structured output, web search and a code interpreter make it a strong workhorse for long-horizon coding and multi-stage agent tasks.
- Text + image + video
- 1M context
- Tools and code
Web search, code interpretation and function calling are route capabilities, not an automatic hosted workspace. Your application must configure them, validate returned sources and control permissions.
AI Chat
Send a prompt and see the response here.
qwen3.8-maxStart a conversation with this model and ask follow-up questions.
USD /1M tokens
Pricing
| Group | Input | Output | Cache Read | Cache Write |
|---|---|---|---|---|
| default | $1.775 | $5.32997 | $0.176968 | $2.219993 |
| qwen-0.8 | $1.42 | $4.263976 | $0.141574 | $1.775994 |
Public list prices are shown in USD. Final charges may vary by account group and usage tier.
Overview
What is Qwen 3.8 Max?
Use Qwen 3.8 Max when a task combines several modalities with a long reasoning chain: a software change with screen evidence, a video-backed report or a multi-stage professional workflow.
The flagship tier is designed for quality, not the lowest-cost batch call. Keep a lighter Qwen route for routine transformations and make escalation a measured routing decision.
Write the decision contract
Name the output, evidence threshold and reviewer. Use structured sections so a long answer remains actionable.
Connect artifacts
Pair screenshots or clips with logs and code, then request a single next check before a broad patch.
Ground every action
Use search and function tools through explicit schemas. Keep source freshness and authorization in the application.
From brief to deliverable
Put Qwen 3.8 Max to work
Product and engineering leads
Plan a long-horizon project
Turn a large product brief, architecture notes and constraints into milestones with dependencies and acceptance checks.
- Tools you need
- Document retrieval, issue tracking and a test environment.
- Keep in mind
- The model can organize plans but cannot make an unapproved commitment or establish a missing dependency.
View a task brief
Use the supplied a dated brief, repository map, constraints and delivery goals. to plan plan a long-horizon project. Return a staged project plan with open decisions and tests., state the evidence limits and list the checks required before accepting the result. Do not claim an output or interaction was verified without direct evidence.
SRE and operations teams
Analyze a video-backed incident
Use selected frames, a transcript and logs to build a diagnosis that separates visible events from inferred causes.
- Tools you need
- Media extraction, log search and a safe reproduction environment.
- Keep in mind
- A recording cannot prove hidden service state or a live fix.
View a task brief
Use the supplied a recording, selected frames, transcript, logs and incident policy. to plan analyze a video-backed incident. Return a source-linked incident brief and verification plan., state the evidence limits and list the checks required before accepting the result. Do not claim an output or interaction was verified without direct evidence.
Research and strategy teams
Run a grounded research workflow
Ask the model to search supplied sources, compare claims and return a structured recommendation with source metadata.
- Tools you need
- Search or retrieval tools, source storage and a human reviewer.
- Keep in mind
- Search results require freshness and source checks; never treat a tool response as self-authenticating.
View a task brief
Use the supplied research questions, allowed sources, freshness rules and output schema. to plan run a grounded research workflow. Return a cited comparison memo and unresolved questions., state the evidence limits and list the checks required before accepting the result. Do not claim an output or interaction was verified without direct evidence.
Capabilities in context
Key features
| Feature | What you get | Why it matters |
|---|---|---|
| Context capacity | 1M tokens and up to 128K output listed in the catalog description | Set an output budget that matches the artifact; a maximum is not a requirement to produce a huge response. |
| Input modes | Text, image and video inputs are listed | Use a source manifest and validate media size, duration and encoding before sending. |
| Agent features | Function calling, structured output, web search and code interpreter are listed | Build schema validation, permissions and source freshness checks around each tool. |
| Flagship positioning | Quality-first workhorse for complex tasks | Compare against a lighter Qwen route on accepted output, latency and cost before making it a default. |
| Version identity | TokenHot model ID: qwen3.8-max | Pin the route in logs and include tool configuration in reproducibility records. |
Capabilities and output limits are taken from TokenHot's current catalog description. Verify tool availability, media constraints and effective output limits on the selected endpoint.
Audience & fit
Where Qwen 3.8 Max fits
Teams shipping complex software
Use the route to connect code, visual evidence and multi-step plans when a missed dependency is expensive.
SRE and operations teams
Keep timestamps, logs and visual evidence together, then reproduce the proposed next check safely.
Research and strategy teams
Use grounded tools with freshness rules and a reviewer for decisions with consequential impact.
Cost context
Plan the cost of a useful result
Budget tools and output
Search, code execution, validation and long outputs can materially change the total. Measure the complete workflow.
Keep source freshness explicit
Search and retrieval results change over time. Store URLs, timestamps and the exact query beside the generated memo.
Compare hard and easy tasks
A flagship route may be wasteful for routine calls. Evaluate a mixed fixture set before changing default routing.
Use the current pricing table for available rates and account groups. Model IDs and prices come from TokenHot’s catalog; provider pricing and subscriptions are separate.
API
Code examples
curl --location --request POST 'https://api.tokenhot.ai/v1/chat/completions' \
--header 'Authorization: Bearer <token>' \
--header 'Content-Type: application/json' \
--data-raw '{
"model": "qwen3.8-max",
"messages": [
{
"role": "user",
"content": "What model are you"
}
]
}'Blog
Related Articles
FAQ
Frequently asked questions
What is Qwen 3.8 Max best suited for?
Qwen 3.8 Max is suited to the input, output, and capability types listed on this page. Test production workloads before choosing it for a critical system.
How is Qwen 3.8 Max priced?
Pricing is shown in the pricing section above. Actual cost depends on usage volume, account group, and the request payload.
How can I call Qwen 3.8 Max?
Use the model ID shown above with one of the supported API protocols. Code examples are provided when usage data is available.
Does Qwen 3.8 Max browse the web automatically?
No. Web search is listed as a capability, but your endpoint and application must configure it and validate the returned sources.
Can it generate video?
The catalog lists video input and text output. Use a video-generation model for a video deliverable.
Is the 128K output limit guaranteed on every route?
It appears in the catalog description. Effective output budgets can depend on the endpoint and request, so verify them in integration.
When should I use a lighter Qwen model?
Use the same acceptance checks on routine and complex tasks. Choose a lighter route when it passes at lower total cost and latency.
Get started
Build with Qwen 3.8 Max
API