Skip to model information

Qwen3.8-2.4T-A95B

Weights available

Open-weight Qwen model with a 256K-token context window. Released August 12, 2026. Weights on Hugging Face.

Access details: Qwen open weights (Hugging Face)

The Hugging Face repository. Download the weights and run them yourself or with a host.

Qwen/Qwen3.8-2.4T-A95B

Also on the Qwen API as qwen3.8-2.4t-a95b

Updated Oct 1, 2026

Capabilities & limits

Context window
256K
Maximum output
Not published
Image input
Unsupported
Tool calling
Not published
Structured output
Not published

Hosted price

USD / 1M tokens
Input
$2.00
Output
$6.00

What Qwen charges for qwen3.8-2.4t-a95b, its API version of this model.

Free to download. Running the weights costs whatever your hardware or host charges.

Price history
Oct 1, 2026

Input Not applicable per million tokens

Oct 1, 2026

Output Not applicable per million tokens

Technical details

Parameters
2.4T
Active parameters
95B
Architecture
Causal Language Model
Training / knowledge cutoff
Not published
Input modalities
Text
Output modalities
Text

License

Other: qwen3.8-max (as declared in the Hugging Face model card metadata)

How they describe it

For the first time, Qwen3.8 brings a Qwen-Max-class model to open release.

Release & lifecycle

Published
Aug 12, 2026
available
As of Oct 1, 2026

Considering a switch?

Specs only go so far. Try both on your own tasks and compare quality, speed and cost.

Compare with GPT-6 Sol →

Prices and availability as of October 20, 2026. Qwen's own pages have the latest.