Qwen4-Max-Preview
A time-limited multimodal reasoning preview for interactive coding and agent tools through Alibaba Cloud Token Plan.
A multimodal reasoning model for applications that combine text, visual understanding, search and tool calling.
Visit website
Qwen4-Max is a multimodal model available through Alibaba Cloud Model Studio. It accepts text, images and video and returns text, serving developers building coding assistants, office applications and agents.
Developers can combine visual questions with text generation, connect application functions through tool calling and request structured responses for downstream processing.
Process image and video inputs alongside text instructions for visual analysis.
A 1,000,000-token context window supports extensive inputs, with deep thinking available for reasoning tasks.
Web search and parallel function calling let applications combine model responses with external information and application functions.
JSON Object and JSON Schema output options support applications that need machine-readable results.
Integrate through OpenAI Chat Completions, OpenAI Responses, Anthropic Messages or DashScope.
Explore a coding assistant that examines supplied code and proposes changes for review.
Build a document assistant that answers questions about supplied text and visual material.
Connect business functions and request structured results for a controlled application workflow.
Application API access defaults to pay-as-you-go billing with separate input, output and caching charges. Token Plan Personal Edition is a separate Credits-based subscription available in Singapore for interactive use within compatible coding and agent tools. It excludes automated scripts, application backends and non-interactive batch calls. See official pricing for current rates.
Check current pricing ↗Use qwen4-max in the model parameter.
Monthly plan Credits are consumed first, followed by available usage-pack Credits. Service pauses when both are exhausted.