Alibaba (Qwen)

Qwen3-235B-A22B-Thinking-2507

Generally availableOpen weights262K contextReasoning

Summary

Qwen3-235B-A22B-Thinking-2507 is an open-weight language model from Alibaba (Qwen), released on 25 Jul 2025. It accepts text and generates text, with a 262K-token context window and up to 82K output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.30 per million input tokens and $3 per million output tokens (Hugging Face Inference Providers), across 2 providers we track. It is scheduled for retirement on 9 Oct 2026.

Qwen reasoning model for deliberate problem solving, math, and coding

Last verified 18 Nov 2026. Every figure on this page links to its source.

Specifications

DeveloperAlibaba (Qwen)
Release date25 Jul 2025 [models.dev]
StatusGenerally available [models.dev]
WeightsOpen weights [Hugging Face]
Licenseapache-2.0 [Hugging Face]
Parameters235B [Hugging Face]
Context window262K (262,144 tokens) [models.dev]
Max output82K tokens [models.dev]
Knowledge cutoffApr 2025 [models.dev]
InputText [models.dev]
OutputText [models.dev]
Reasoning modeYes [models.dev]
Tool callingYes [models.dev]
Structured outputNo [OpenRouter]
Retirement date9 Oct 2026 [OpenRouter]
Hugging FaceQwen/Qwen3-235B-A22B-Thinking-2507 [Hugging Face]

Qwen3-235B-A22B-Thinking-2507 API pricing by provider

USD per million tokens. Sorted by input price.

ProviderInputOutputCached inputContextSource
Hugging Face Inference Providers
Qwen/Qwen3-235B-A22B-Thinking-2507
$0.30$3–262Kmodels.dev, 18 Nov 2026
OpenRouter
qwen/qwen3-235b-a22b-thinking-2507
$0.45$3.50–128KOpenRouter, 18 Nov 2026

Running Qwen3-235B-A22B-Thinking-2507 locally

Estimated memory, from the parameter count and the architecture in config.json on Hugging Face. Includes about 0.8 GB of runtime overhead.

QuantizationWeightsTotal, 8K contextTotal, 32K context
Q4_K_M (4-bit)143 GB145 GB150 GB
Q8_0 (8-bit)250 GB252 GB257 GB
FP16 / BF16 (16-bit)470 GB473 GB477 GB

At 4-bit with an 8K conversation, Qwen3-235B-A22B-Thinking-2507 needs about 145 GB, more than any single consumer graphics card; it needs a unified-memory machine, a multi-GPU setup or a split between graphics card and system RAM. Check your own hardware, or see Best computer for local AI on our sister site AIPCs.

Frequently asked questions

What is the context window of Qwen3-235B-A22B-Thinking-2507?

Qwen3-235B-A22B-Thinking-2507 has a context window of 262,144 tokens (262K) and can generate up to 81,920 tokens in a single response, according to models.dev.

How much does the Qwen3-235B-A22B-Thinking-2507 API cost?

Qwen3-235B-A22B-Thinking-2507 is listed from $0.30 per million input tokens and $3 per million output tokens (Hugging Face Inference Providers). Prices last verified 18 Nov 2026.

Is Qwen3-235B-A22B-Thinking-2507 open source?

Qwen3-235B-A22B-Thinking-2507 is an open-weight model: its weights are published on Hugging Face as Qwen/Qwen3-235B-A22B-Thinking-2507 under the apache-2.0 license. Check the license terms before commercial use.

When was Qwen3-235B-A22B-Thinking-2507 released?

Alibaba (Qwen) released Qwen3-235B-A22B-Thinking-2507 on 25 Jul 2025, according to models.dev.

Which providers offer Qwen3-235B-A22B-Thinking-2507?

We track Qwen3-235B-A22B-Thinking-2507 on 2 providers: Hugging Face Inference Providers, OpenRouter.

How much VRAM does Qwen3-235B-A22B-Thinking-2507 need?

About 145 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 252 GB at 8-bit. That is more than any single consumer graphics card holds. These are estimates from the model's published parameter count and architecture.

Similar models from other labs

Comparable on output type, weights, release date, price and context window. Prices are the lowest listed, USD per million tokens.

ModelLabReleasedContextInput $/1MOutput $/1MWeights
Kimi-K2-Instruct-0905Moonshot AI5 Sept 2025262K$0.60$2.50Open
Kimi-K2-ThinkingMoonshot AI6 Nov 2025262K$0.60$2.50Open
GLM-4.5Zhipu AI (Z.ai)28 Jul 2025131K$0.60$2.20Open
GLM-4.6Zhipu AI (Z.ai)30 Sept 2025205K$0.50$2Open
DeepSeek-R1-0528DeepSeek28 May 2025164K$0.50$2.15Open
Hermes-4-405BNousresearch26 Aug 2025131K$1$3Open

Change history

  • Name changed from Qwen3 235B A22B Thinking 2507 to Qwen3-235B-A22B-Thinking-2507
  • Context window changed from 131K to 262K
  • Max output changed from 118K to 82K
  • Knowledge cutoff changed from 30 Jun 2025 to Apr 2025
  • Now listed on Hugging Face Inference Providers
  • OpenRouter price changed: input $0.23 → $0.45, output $2.30 → $3.50 per 1M tokens

Sources