Qwen3 32B

Single-GPU Qwen3 with reasoning; competitive with much larger models.

Capabilities

Modalities: text→text
Capabilities: reasoningbatchSupportpromptCachingfunctionCallingstructuredOutputs
Strengths: Single-GPU capable, Good reasoning for size, Apache 2.0
Tradeoffs: Text-only, lower ceiling than 235B

Official Links

Intro

Docs

Pricing

Release note

Benchmark	Version	Score	Date	Source	Notes

Release	Alias	Lifecycle	Release Date	Deprecation	Shutdown	Summary
Qwen3 32B	qwen3-32b	Active	Apr 29, 2025	—	—	Current published model family snapshot.

Host	Type	Context	Pricing Note	Differences
Alibaba DashScope	first-party	131.1K	$0.14/$0.56 per MTok.	—

Migration Guidance

Best single-GPU open reasoning model in this class.

Replacement models: qwen3-235b-a22b

Change Events

Date	Type	Title	Description	Source
Apr 29, 2025	family_added	Qwen3 32B published	Initial public model family launch.	Qwen3 32B release notes