Qwen3.5 122B A10B
Neural Network
Qwen3.5 122B A10B is a native Qwen vision-language model for text and images, available via API.
Specifications
- Input
- text, images, documents
- Output
- text
- Context size
- 262 144 tokens
- Max answer length
- 65 536 tokens
- Reasoning effort
- low, medium, high
- Capabilities
- reasoning, function calling
Prompt cost
(per 1M tokens)
Answer cost
(per 1M tokens)
How it works Qwen3.5 122B A10B?
Qwen3.5 122B A10B is a native vision-language model from the Alibaba Qwen family that accepts text and images in a single request. Its hybrid architecture combines linear attention with a sparse mixture of experts: out of 122 billion parameters, about 10 billion are activated per token, making inference more efficient. In BotHub, it is accessible from Russia without a VPN or foreign card, and you pay per token with no subscription or expiring balance. Over 250 neural networks are available in the same window, allowing you to compare responses for a single prompt. A unified OpenAI-compatible API lets you switch models in your code with one line. Chats are encrypted via AES-GCM, and for companies, we offer contracts, invoices, and an admin panel with limits. Use cases are simple: analyze a screenshot or diagram and explain what is happening, extract data from a document photo, describe image content for a catalog, build a multimodal assistant via API, or integrate the model into a service where cost-effective inference is crucial.Frequently asked questions about Qwen3.5 122B A10B
You can use generated results for commercial purposes. You own all rights to the content you create. The only restriction: make sure your prompt does not include copyrighted third-party material. You are responsible for respecting the rights to any input data.
This is a text model from Qwen with a 262,144 token context. It is suitable for working with code (writing, refactoring, debugging), analyzing large documents, STEM tasks, and analytics. You can input text, documents, and images, and receive a detailed analysis or ready-to-use code.
The output limit is up to 65,536 tokens per response. This is enough for a large code module, detailed technical documentation, a comprehensive analytical report, or translating long text without splitting the task into parts and manually stitching fragments together.
Yes, reasoning mode is supported: the model breaks down the task step-by-step before providing a result. This significantly helps with math, logic, STEM tasks, and code debugging, where it is important to follow the chain of thought rather than guessing the result immediately.
Yes, the model works with function calling and structured JSON output. This is convenient for agents, document parsing, and integrations with your services. A unified OpenAI-compatible API is available in BotHub, so connecting will not require rewriting code.
The model accepts images and documents: you can send a screenshot, diagram, table, or PDF and ask to analyze the content. Audio and video input is not noted in our data. Image, audio, and video generation in BotHub is available through separate models in the same window.
The Qwen family of models was trained on multilingual data and works confidently with Russian-language queries. Practical tip: compare the response with another model right in BotHub — switching takes a second, and you pay per token with no subscription.