LLM·Dex
Newsletter
Open weightsAlibabaopentextvision

Qwen2-VL-72B

Top open-weight vision-language model, strong on document understanding and chart analysis.

Updated

Quick facts

Released
Aug 2024
Context
128K tokens
Output / 1M
Pricing not published
License
Apache-2.0

About Qwen2-VL-72B

Qwen2-VL-72B is the best open-weight model for vision-language tasks, particularly strong on document AI and Chinese-language image understanding. Apache-2.0 licensed and broadly deployable across vLLM, SGLang, and HuggingFace's Inference Endpoints.

For self-hosted document-extraction pipelines, Qwen2-VL is frequently the top pick. The 72B size class is heavy but the quality justifies the inference cost on serious workloads.

Benchmarks

Published scores from Alibaba's model card or independent leaderboards. We do not publish numbers we cannot source, see methodology.

HumanEval
,
Python coding pass@1
MMLU
,
Broad academic knowledge
GPQA
,
Graduate-level reasoning
SWE-bench
,
Real software-engineering tasks
Benchmark scores not yet available. We only publish numbers we can source from official model cards or independent leaderboards, see methodology.

Capabilities

Strengths

  • Top open vision-language model
  • Apache-2.0
  • Strong on documents

Tracked weaknesses

  • Heavy at 72B

Pricing

Per-million-token rates as published by Alibaba.

Per-token pricing not yet published for Qwen2-VL-72B. Check the official provider site for current tiers.

Call Qwen2-VL-72B from your code

Drop-in snippet for the Alibaba SDK. Set your API key in the environment and run.

typescript
import OpenAI from "openai";

const client = new OpenAI({
  // Use OPENAI_API_KEY for OpenAI, or your provider's key + baseURL.
  apiKey: process.env.OPENAI_API_KEY,
});

const completion = await client.chat.completions.create({
  model: "qwen-2-vl-72b",
  messages: [
    { role: "user", content: "What's the time complexity of quicksort?" },
  ],
});

console.log(completion.choices[0].message.content);

Best for

Tasks where Qwen2-VL-72B ranks among LLMDex's top picks.

Compare Qwen2-VL-72B with…

Frequently asked

  • How much does Qwen2-VL-72B cost?
    Alibaba has not published per-token API pricing for Qwen2-VL-72B at the time of writing. Check the official site for current pricing tiers, or compare against alternative models on LLMDex.
  • What is Qwen2-VL-72B's context window?
    Qwen2-VL-72B supports a context window of 128K tokens.
  • Is Qwen2-VL-72B open source?
    Qwen2-VL-72B ships with open weights under the Apache-2.0 license. You can self-host it, fine-tune it, and (subject to the license terms) deploy it commercially.
  • When was Qwen2-VL-72B released?
    Qwen2-VL-72B was released on Aug 29, 2024 by Alibaba.
Friday digest

Intelligence, distilled weekly.

One short email every Friday, new model launches, leaderboard moves, and pricing drops. Curated by hand. Free, no spam.