Qwen3 30B A3B Instruct 2507
by Prime Inference · family qwen · listed Jul 2025TW
Updated compact Qwen3 MoE instruct model for multilingual chat, coding, and tool use
Current rates USD per 1M tokens
Input$0.13
Output$0.52
Cache read—
Cache write—
Against the market
On the record
- Model id
qwen/qwen3-30b-a3b-instruct-2507- Context window
- 262K tokens
- Max output
- 236K tokens
- Input modalities
- text
- Output modalities
- text
- Reasoning
- no
- Tool calls
- yes
- Open weights
- yes
- Knowledge cutoff
- —
- Released
- 2025-07-28
- Last updated
- 2025-07-28
- Provider docs
- docs.primeintellect.ai/inference/overview
Back of the envelope
≈
Also listed at
- ModelScopefree in · free out
- Ambient$0.09 in · $0.30 out
- SiliconFlow$0.09 in · $0.30 out
- Cortecs$0.099 in · $0.299 out
- Kilo Gateway$0.10 in · $0.30 out
- Eden AI$0.10 in · $0.30 out
- Nebius Token Factory$0.10 in · $0.30 out
- OpenRouter$0.10 in · $0.30 out
- SiliconFlow (China)$0.105 in · $0.418 out
- Qiniu$0.109 in · $0.435 out
- NanoGPT$0.20 in · $0.50 out
- Clarifai$0.30 in · $0.50 out