Qwen3Open Weight
Released
Qwen3 is an AI model released by Qwen on Tuesday, Apr 29 2025, 54 days after QwQ-32B. It is an open-weight model — the trained weights are available to download and run. It is a 235B parameter model with a 128k token context window.
Get Qwen releases and AI news.
One email, every other Monday.
API pricing
- InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
- $0.70
- OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
- $2.80
- InputWhat you pay for everything you send the model — your question, plus any documents or earlier conversation you include with it.
- $0.70
- OutputWhat you pay for the text the model writes back. It is normally the dearer half: producing an answer costs more than reading one.
- $8.40
Available from
| StreamLake | 30b-a3b-instruct-2507 | — | $0.0482 | $0.1931 | 128K | — |
|---|---|---|---|---|---|---|
| NextBit | 14b | — | $0.10 | $0.22 | 41K | int4 |
| DeepInfra | 14b | — | $0.12 | $0.24 | 41K | fp8 |
| DeepInfra | 32b | — | $0.08 | $0.28 | 41K | fp8 |
| DekaLLM | 30b-a3b-instruct-2507 | — | $0.09 | $0.30 | 256K | — |
| SiliconFlow | 30b-a3b-instruct-2507 | — | $0.09 | $0.30 | 256K | fp8 |
| Nebius | 30b-a3b-instruct-2507 | — | $0.10 | $0.30 | 256K | fp8 |
| GMICloud | 235b-a22b-2507 | — | $0.0875 | $0.35 | 256K | fp8 |
| DeepInfra | 30b-a3b | — | $0.12 | $0.50 | 41K | fp8 |
| DeepInfra | 235b-a22b-2507 | — | $0.09 | $0.55 | 256K | fp8 |
| SiliconFlow | 32b | — | $0.14 | $0.57 | 128K | fp8 |
| Novita | 235b-a22b-2507 | — | $0.09 | $0.58 | 128K | fp8 |
| Nebius | 235b-a22b-2507 | — | $0.20 | $0.60 | 256K | fp8 |
| Venice | 235b-a22b-2507 | — | $0.15 | $0.75 | 128K | fp8 |
| Parasail | 235b-a22b-2507 | — | $0.14 | $0.80 | 128K | fp8 |
| StreamLake | 235b-a22b-2507 | — | $0.21 | $0.84 | 128K | — |
| 235b-a22b-2507 | us-south1 | $0.25 | $1.00 | 256K | — | |
| Novita | 235b-a22b-thinking-2507 | — | $0.30 | $3.00 | 128K | fp8 |
| Venice | 235b-a22b-thinking-2507 | — | $0.45 | $3.50 | 128K | fp8 |
Compare Qwen3 with
About
Qwen3, released April 29, 2025, unified two things Alibaba had previously shipped separately: a conventional instruct model and a reasoning model. Every Qwen3 model could switch between thinking and non-thinking mode within a single set of weights, with the thinking budget controllable per request — the same hybrid idea Anthropic had introduced with Claude 3.7 Sonnet two months earlier, but in open weights.
The family spanned dense models at 0.6B, 1.7B, 4B, 8B, 14B and 32B plus two mixture-of-experts models, 30B-A3B and the 235B-A22B flagship, all trained on 36 trillion tokens across 119 languages and dialects. Crucially, all of it went out under Apache 2.0 including the flagship — the first time Qwen had put the largest model of a generation under a fully permissive licence, and a direct contrast with Meta's custom Llama terms.
Frequently asked questions
Qwen3 was released by Qwen on Tuesday, Apr 29 2025.