Alibaba (Qwen)

Qwen2.5-Coder-32B-Instruct

Generally availableOpen weights131K context

Summary

Qwen2.5-Coder-32B-Instruct is an open-weight language model from Alibaba (Qwen), released on 12 Nov 2024. It accepts text and generates text, with a 131K-token context window and up to 8K output tokens per response. As of 18 Nov 2026, the lowest listed API price is $0.06 per million input tokens and $0.20 per million output tokens (Hugging Face Inference Providers), across 3 providers we track.

Qwen coding model for software agents, repository edits, and code reasoning

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

Specifications

DeveloperAlibaba (Qwen)
Release date12 Nov 2024 [models.dev]
StatusGenerally available [models.dev]
WeightsOpen weights [Hugging Face]
Licenseapache-2.0 [Hugging Face]
Parameters32.8B [Hugging Face]
Context window131K (131,072 tokens) [models.dev]
Max output8K tokens [models.dev]
Knowledge cutoffApr 2024 [models.dev]
InputText [models.dev]
OutputText [models.dev]
Reasoning modeNo [models.dev]
Tool callingYes [models.dev]
Structured outputYes [models.dev]
Hugging FaceQwen/Qwen2.5-Coder-32B-Instruct [Hugging Face]

Qwen2.5-Coder-32B-Instruct API pricing by provider

USD per million tokens. Sorted by input price.

ProviderInputOutputCached inputContextSource
Hugging Face Inference Providers
Qwen/Qwen2.5-Coder-32B-Instruct
$0.06$0.20–131Kmodels.dev, 18 Nov 2026
Cloudflare Workers AI
@cf/qwen/qwen2.5-coder-32b-instruct
$0.66$1–33Kmodels.dev, 18 Nov 2026
OpenRouter
qwen/qwen-2.5-coder-32b-instruct
$0.66$1–33KOpenRouter, 18 Nov 2026

Running Qwen2.5-Coder-32B-Instruct 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)20 GB23 GB29 GB
Q8_0 (8-bit)35 GB38 GB44 GB
FP16 / BF16 (16-bit)66 GB68 GB75 GB

At 4-bit with an 8K conversation, Qwen2.5-Coder-32B-Instruct needs about 23 GB, so it fits on a graphics card with 24 GB or more, such as the NVIDIA GeForce RTX 3090 24 GB. Check your own hardware, or see Best GPU for local LLMs on our sister site AIPCs.

Frequently asked questions

What is the context window of Qwen2.5-Coder-32B-Instruct?

Qwen2.5-Coder-32B-Instruct has a context window of 131,072 tokens (131K) and can generate up to 8,192 tokens in a single response, according to models.dev.

How much does the Qwen2.5-Coder-32B-Instruct API cost?

Qwen2.5-Coder-32B-Instruct is listed from $0.06 per million input tokens and $0.20 per million output tokens (Hugging Face Inference Providers). Prices last verified 18 Nov 2026.

Is Qwen2.5-Coder-32B-Instruct open source?

Qwen2.5-Coder-32B-Instruct is an open-weight model: its weights are published on Hugging Face as Qwen/Qwen2.5-Coder-32B-Instruct under the apache-2.0 license. Check the license terms before commercial use.

When was Qwen2.5-Coder-32B-Instruct released?

Alibaba (Qwen) released Qwen2.5-Coder-32B-Instruct on 12 Nov 2024, according to models.dev.

Which providers offer Qwen2.5-Coder-32B-Instruct?

We track Qwen2.5-Coder-32B-Instruct on 3 providers: Hugging Face Inference Providers, Cloudflare Workers AI, OpenRouter.

How much VRAM does Qwen2.5-Coder-32B-Instruct need?

About 23 GB at 4-bit (Q4_K_M) with an 8K-token conversation, and about 38 GB at 8-bit. The 4-bit model fits on a graphics card with 24 GB or more, such as the NVIDIA GeForce RTX 3090 24 GB. 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
Command R7BCohere13 Dec 2024128K$0.04$0.15Open
Phi-4-miniMicrosoft11 Dec 2024128K$0.07$0.30Open
Llama-3.3-70B-InstructMeta6 Dec 2024128K$0.10$0.32Open
Llama 3.2 3B InstructMeta25 Sept 2024131K$0.05$0.34Open
Gemma 3 27B ITGoogle DeepMind12 Mar 2025131K$0.08$0.16Open
Command R7B ArabicCohere27 Feb 2025128K$0.04$0.15Open

Change history

  • Context window changed from 33K to 131K
  • Max output changed from 29K to 8K
  • Structured output changed from no to yes
  • Tool calling changed from no to yes
  • Max output changed from 33K to 29K
  • Name changed from Qwen2.5 Coder 32B Instruct to Qwen2.5-Coder-32B-Instruct
  • Structured output changed from yes to no
  • Context window changed from 131K to 33K
  • Tool calling changed from yes to no
  • Max output changed from 8K to 33K
  • Name changed from Qwen2.5-Coder-32B-Instruct to Qwen2.5 Coder 32B Instruct
  • Max output changed from 33K to 8K
  • Name changed from Qwen2.5 Coder 32B Instruct to Qwen2.5-Coder-32B-Instruct
  • Structured output changed from no to yes
  • Tool calling changed from no to yes
  • Context window changed from 33K to 131K
  • Name changed from Qwen2.5-Coder-32B-Instruct to Qwen2.5 Coder 32B Instruct
  • Max output changed from 29K to 33K
  • Max output changed from 8K to 29K
  • Tool calling changed from yes to no
  • Context window changed from 131K to 33K
  • Structured output changed from yes to no
  • Name changed from Qwen2.5 Coder 32B Instruct to Qwen2.5-Coder-32B-Instruct
  • Tool calling changed from no to yes
  • Structured output changed from no to yes

Sources