---
id: "qwen/qwen3-30b-a3b-instruct-2507"
also: ["qwen/qwen3-30b-a3b-instruct-2507", "Qwen/Qwen3-30B-A3B-Instruct-2507", "requesty/alibaba/qwen3-30b-a3b-instruct-2507"]
cheapest_list_usd_per_m: "0.0481 in, 0.193 out, at streamlake on openrouter"
cheapest_effective_usd_per_m: "0.0479 in, 0.1897 out, at streamlake on openrouter"
context: 262144
max_output: 235929
input: ["text"]
output: ["text"]
released: "2025-07-29"
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["frequency_penalty", "logit_bias", "logprobs", "max_tokens", "presence_penalty", "repetition_penalty", "response_format", "seed", "stop", "structured_outputs", "temperature", "tool_choice", "tools", "top_k", "top_logprobs", "top_p"]
endpoints: 8
gateways: "openrouter (6), requesty (2)"
no_training: 8
trains: 0
retains_prompts: 2
---
# Qwen: Qwen3 30B A3B Instruct 2507

Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and agentic tool use. Post-trained on instruction data, it demonstrates competitive performance across reasoning (AIME, ZebraLogic), coding (MultiPL-E, LiveCodeBench), and alignment (IFEval, WritingBench) benchmarks. It outperforms its non-instruct variant on subjective and open-ended tasks while retaining strong factual and coding performance.