Qwen3.5 397B
Qwen3.5 397B is a multimodal text generation model from Qwen with a 262,144-token context window.
Qwen
DeepInfraLarge-scale reasoning with multimodal input
Qwen3.5 397B is a large mixture-of-experts language model developed by Qwen, released in February 2026. It carries 397 billion total parameters while activating approximately 17 billion per forward pass, a design that allows it to handle complex tasks without proportionally scaling compute at inference time. The model supports both text and image inputs, making it capable of processing multimodal prompts within a single conversation turn.
With a context window of 262,144 tokens, Qwen3.5 397B is suited for tasks that require processing long documents, extended conversations, or large codebases in a single pass. Its reasoning tag indicates it is designed to handle multi-step logical problems, making it applicable to tasks such as mathematical reasoning, code generation, and structured analysis. The model is served through DeepInfra on MindStudio, requiring no API key management from the user.
What Qwen3.5 397B supports
Extended Context Window
Processes up to 262,144 tokens in a single pass, enabling analysis of long documents, large codebases, or extended conversation histories without truncation.
Multi-Step Reasoning
Handles complex logical and mathematical problems by breaking them into sequential steps, as indicated by the model's REASONING tag.
Image Understanding
Accepts image inputs alongside text prompts, supporting tasks like visual question answering and image-based analysis within a single request.
Mixture-of-Experts Architecture
Uses 397 billion total parameters but activates only approximately 17 billion per forward pass, reducing per-token compute while retaining large model capacity.
Code Generation
Generates, explains, and debugs code across common programming languages, supported by the model's large parameter count and reasoning capabilities.
Ready to build with Qwen3.5 397B?
Get Started FreeCommon questions about Qwen3.5 397B
What is the context window size for Qwen3.5 397B?
Qwen3.5 397B has a context window of 262,144 tokens, which is also the maximum response size. This allows very long documents or conversations to be processed in a single request.
Does Qwen3.5 397B support image inputs?
Yes. The model supports image inputs in addition to text, as indicated by its MULTI-MODAL tag and the supportsImages field in its metadata.
How many parameters does Qwen3.5 397B have, and how many are active at once?
The model has 397 billion total parameters. Its mixture-of-experts design activates approximately 17 billion parameters per forward pass, reflected in the model's full name Qwen3.5-397B-A17B.
When was Qwen3.5 397B released?
Qwen3.5 397B was released in February 2026 according to the model metadata.
What is the pricing for Qwen3.5 397B on MindStudio?
Pricing information is not published in the current metadata. You can check MindStudio's model catalog or DeepInfra's pricing page for the latest per-token rates.
Parameters & options
Explore similar models
QwenQwen3.8 2.4T
Qwen3.8 2.4T is a mixture-of-experts text generation model from Qwen with a 262,144 token context window.
QwenQwen3.6 Flash
Qwen3.6 Flash is a text and image understanding model from Qwen with a 1,000,000-token context window and optional thinking mode.
QwenQwen3.7 Max
Qwen3.7 Max is a text generation model from Qwen with a 1,000,000-token context window and optional deep thinking mode.
AnthropicClaude 5 Sonnet
Claude 5 Sonnet is Anthropic's flagship text and vision model with a 1,000,000-token context window and extended reasoning.
OpenAIGPT-5.5
GPT-5.5 is a multimodal text generation model from OpenAI with a 1,000,000-token context window and adjustable reasoning effort.
GoogleGemma 4 31B
Gemma 4 31B is a multimodal text generation model from Google with a 262,144-token context window.
Start building with Qwen3.5 397B
No API keys required. Create AI-powered workflows with Qwen3.5 397B in minutes — free.