Alibaba: qwen3-235b-a22b API
Alibaba Cloud's flagship MoE model released on April 29, 2025, with a significant update (version 2507) on July 21, 2025. It features 235 billion total parameters, activating only 22 billion (22B) per token (hence A22B). Described as having "22B speed with 235B intelligence," it represents the pinnacle of the Qwen3 series. Supporting native Dual-Mode, it rivals GPT-4o and Claude 4 in complex reasoning (AIME25), competitive programming, and ultra-long document analysis.
- Context window: 131,072 tokens
- Max output: 16,384 tokens
- Input: text
- Output: text
- Reasoning: Supported
- Tool calling: Supported
- Released: 2025-03-31
- Knowledge cutoff: 2025-05-31
Frequently Asked Questions
What is the context window of qwen3-235b-a22b?
qwen3-235b-a22b supports a context window of up to 131,072 tokens.
Does qwen3-235b-a22b support function calling?
Yes. qwen3-235b-a22b supports tool / function calling.
Does qwen3-235b-a22b support reasoning?
Yes. qwen3-235b-a22b is a reasoning-capable model.
What is the knowledge cutoff of qwen3-235b-a22b?
The knowledge cutoff of qwen3-235b-a22b is 2025-05-31.