Qwen3 32B — pricing & specs
Qwen3 32B is a balanced model from Alibaba, at capability tier 3 of 8. Live $/M-token pricing and specs from the Mintok catalog, as of 2026-06-13.
$/M-token pricing
as of 2026-06-13 · OpenRouter (lowest)What Qwen3 32B costs in practice
At $0.28/M output tokens, Qwen3 32B is the 2nd-cheapest of 9 priced tier-3 models in the Mintok catalog — about 53% below the tier median of $0.60/M.
| Scenario | Est. cost |
|---|---|
| 1M input + 1M output tokens | $0.360 |
| A 10K-token request (8K in / 2K out) | $0.0012 |
| 1,000 agent runs (~2K in / 500 out each) | $0.300 |
| 1M-token RAG chat session (900K cached in / 100K out) | $0.100 |
Estimated from list price (as of 2026-06-13) at the token mixes shown; your real cost depends on prompt/response length and cache hit-rate. Model your own volume →
Specifications
- Vendor
- Alibaba
- Family
- Qwen3
- Capability tier
- 3 / 8
- Size class
- BALANCED
- Category
- GENERAL
- Parameters
- 32B total · 32B active (MoE)
- Context window
- 131,072 tokens
- Licensing
- Proprietary
Carbon footprint (modeled)
Modeled decode on a reference Cerebras Cerebras WSE-3 Turbo at 50% MBU, batch 32, global-avg grid. Compare all models & grids →
Alternatives at tier 3
Same capability tier, sorted by output price — the cheaper-capable options.
- Mistral NeMoMistral$0.15/M out
- Gemma 4 31BGoogle$0.35/M out
- Gemini 2.5 Flash-LiteGoogle$0.40/M out
- Mistral Small 4Mistral$0.60/M out
- Mixtral 8x7BMistral$0.70/M out
- Magistral SmallMistral$1.50/M out
FAQ
How much does Qwen3 32B cost per token?
Qwen3 32B costs $0.08 per million input tokens and $0.28 per million output tokens, via OpenRouter (lowest), as of 2026-06-13.
Is Qwen3 32B expensive?
At $0.28/M output tokens, Qwen3 32B is the 2nd-cheapest of 9 priced tier-3 models in the Mintok catalog — about 53% below the tier median of $0.60/M.
What does a typical request with Qwen3 32B cost?
At list price, a 10K-token request (8K in / 2K out) with Qwen3 32B costs about $0.0012, and 1,000 agent runs (~2K in / 500 out each) about $0.300.
What is Qwen3 32B's context window?
Qwen3 32B supports a 131,072-token context window.
What capability tier is Qwen3 32B?
Qwen3 32B sits at capability tier 3 of 8 on the Mintok scale (BALANCED size class).
Is Qwen3 32B open-source or proprietary?
Qwen3 32B is a proprietary model from Alibaba, accessed via API.
Put Qwen3 32B to work
Benchmark it on your prompts, or size a deployment and compare $/M-token against alternatives.
Pricing and specs are maintained by Mintok's catalog watchers from public sources. Not investment advice. Methodology.