Qwen2.5 72B Instruct
VerifiedOpen-weight LLM optimized for instruction following and long-context reasoning.
About Qwen2.5 72B Instruct
Built on a transformer architecture, Qwen2.5 72B Instruct was trained on diverse multilingual and code corpora. Its open-weight release enables researchers and developers to inspect, fine-tune, and deploy the model locally or in private environments.
The 131K token context supports extended documents and multi-turn conversations without truncation. Typical applications include coding assistance, technical writing, data summarization, and building domain-specific chat agents.
Users commonly integrate the model via frameworks such as Hugging Face Transformers or vLLM for both research experiments and production inference workloads.
Capabilities
Benchmarks & performance
Independent evaluation scores and measured speed.
Source: Artificial Analysis
How Qwen2.5 72B Instruct compares
Qwen2.5 72B Instruct (striped bar) vs other language models on intelligence, speed and price.
Intelligence
Artificial Analysis Intelligence Index · Higher is better · Qwen2.5 72B Instruct ranks #50 of 67
Price
USD per 1M output tokens · Lower is better · Qwen2.5 72B Instruct ranks #44 of 141
Sources: Artificial Analysis (intelligence, speed) · OpenRouter (price).
Best for
Long-Document Analysis
The model handles inputs up to 131072 tokens for in-depth reasoning over lengthy texts such as research papers, legal contracts, or technical documentation.
Software Development Assistance
It supports code generation, debugging, and tool use to help developers write, test, and refine programs in multiple languages.
Multilingual Math and Instruction Tasks
Strong performance in mathematical problem solving combined with multilingual text generation and instruction following enables accurate responses across languages and complex calculations.
Strengths & limitations
Strengths
- +Strong coding and technical reasoning
- +Robust multilingual support including Chinese
- +Effective long-context handling
- +Clear and structured responses
Limitations
- –Text-only modality
- –No native vision or multimodal input
- –Knowledge cutoff limits real-time information
Cost calculator
Estimate what Qwen2.5 72B Instruct would cost for your usage.
Based on Qwen2.5 72B Instruct's $0.36/1M input · $0.40/1M output. Estimate only — actual cost varies by provider and caching.
Quick start
OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. Only the model slug changes between models.
import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://openrouter.ai/api/v1",
apiKey: process.env.OPENROUTER_API_KEY,
});
const completion = await client.chat.completions.create({
model: "qwen/qwen-2.5-72b-instruct",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(completion.choices[0].message.content);Model slug: qwen/qwen-2.5-72b-instruct
Editor's verdict
Qwen2.5 72B Instruct is Alibaba Qwen's open-weight language models with a 131K-token context window.
On independent testing it scores 15.6 on the Artificial Analysis Intelligence Index.
At $0.40 per 1M output tokens, it is very cost-efficient for its class.
As an open-weight model you can self-host it or call it through a hosted API.
Best suited to strong coding and technical reasoning and robust multilingual support including chinese.
Frequently asked questions
The model provides a context window of 131072 tokens.
User reviews
Real, verified reviews from the community shape this model's rating.
Loading reviews…
Other Qwen models
Sibling versions in the Qwen family from Alibaba Qwen.
Similar models
Other language models worth comparing.