Pass IndexThe State of AISign in

Qwen3 Max Thinking

Compared with the snapshot as of September 23, 2025, the Qwen-3 series Max model in this release achieves an effective integration of thinking and non-thinking modes, resulting in a comprehensive and substantial improvement in the model’s overall performance.

text → text · made by Alibaba

$0.78$3.9per Mtok in / out
Inference.net · 5 sellers

Sold by 5 ways

SellerLaneRate
Inference.net aggregatorstandard$0.78per Mtok in$3.9per Mtok outapi.inference.net · read 2026-09-16
Nous Research aggregatorstandard$0.78per Mtok in$3.9per Mtok outinference-api.nousresearch.com · read 2026-09-12
OpenRouter aggregatorstandard$0.78per Mtok in$3.9per Mtok outopenrouter.ai · read 2026-08-24
DeepInfra aggregatorstandard$1.2per Mtok in$6per Mtok outapi.deepinfra.com · read 2026-08-25
Vercel AI Gateway aggregatorstandard$1.2per Mtok in$6per Mtok out$0.24per Mtok cachedai-gateway.vercel.sh · read 2026-08-25

About

Qwen3 Max Thinking — a text model from Alibaba, sold by 5 companies from $0.78 in and $3.9 out per million tokens.

It takes text and returns text, with a context window of 256,000 tokens. It was published in January 2026. Its sellers say it can reason step by step and call a tool. The catalogue files it under search, extract, sandbox and reasoning. Five companies sell it. The cheapest is $0.78 in and $3.9 out per million tokens at Inference.net.

Every current figure

Maker
Alibaba
Register
model
Takes
text
Returns
text
Context
256,000 tokens
Longest answer
65,536 tokens
Published
January 2026
Licence
not read
Sellers
5
Maker's own price
not read
Price
$0.78 in and $3.9 out per million tokens — Inference.net

Known as 2 names

Qwen/Qwen3-Max-Thinkingalibaba/qwen3-max-thinking