Requesty
Alibaba Cloud logo
Alibaba Cloud

qwen2.5-vl-72b-instruct

Qwen2.5-VL-72B is Alibaba's large vision-language model capable of understanding images and text, with 72B parameters for high-quality visual reasoning and document understanding.

👁Vision🔧Tool calling⚡Caching
Try qwen2.5-vl-72b-instructCompareDocs

Specifications

Context window131K tokens
Max output8K tokens
API typechat
AddedMay 27, 2026
Model IDalibaba/qwen2.5-vl-72b-instruct
Data retentionNo
Used for trainingNo
Provider location🇸🇬 Singapore
Privacy policyAlibaba Cloud Privacy Policy→

Benchmarks

Benchmarks haven't been published yet for this exact variant.

Some variants (region-specific deployments, highspeed tiers) share benchmarks with their base model — check the base model page or the Alibaba Cloud models overview.

Pricing

Input / 1M
$2.80
Output / 1M
$8.40
Cache write / 1M
$3.50
Cache read / 1M
$0.28
Estimated cost
100K input + 10K output$0.36
1M input + 100K output$3.64
10M input + 1M output$36.40

Requesty charges exactly what the upstream provider charges — no markup, no per-request fees. Prompt caching and smart routing can reduce effective cost by 30-80%.

Quickstart

Drop-in compatible with the OpenAI SDK. Change the base URL, swap in your Requesty API key, and set the model to alibaba/qwen2.5-vl-72b-instruct.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
from openai import OpenAI

client = OpenAI(
    api_key="YOUR_REQUESTY_API_KEY",
    base_url="https://router.requesty.ai/v1",
)

response = client.chat.completions.create(
    model="alibaba/qwen2.5-vl-72b-instruct",
    messages=[
        {"role": "user", "content": "Explain quantum computing in one paragraph."},
    ],
)

print(response.choices[0].message.content)

Other Alibaba Cloud models

qwen3.6-plus
1M ctx$0.50/1M
qwen2.5-vl-32b-instruct
131K ctx$1.40/1M
qwen3.7-max
1.0M ctx$2.50/1M
qwen3-max
262K ctx$0.86/1M
qwen3-30b-a3b-instruct-2507
131K ctx$0.20/1M
qwen3-coder-plus
1.0M ctx$1.00/1M

Frequently asked questions

How much does qwen2.5-vl-72b-instruct cost?
qwen2.5-vl-72b-instruct is priced at $2.80 per million input tokens and $8.40 per million output tokens when accessed via Requesty. Prompt caching is supported, which can cut effective input cost by up to 90% on repeated context. Requesty charges exactly what the upstream provider charges — we don't add markup.
What is the context window of qwen2.5-vl-72b-instruct?
qwen2.5-vl-72b-instruct has a context window of 131K tokens, with a maximum output of 8K tokens per response. That's roughly 175 words of input you can fit in a single prompt.
What can qwen2.5-vl-72b-instruct do?
qwen2.5-vl-72b-instruct supports vision input, tool calling, prompt caching. You can call it through any OpenAI-compatible client by pointing base_url to Requesty.
How do I use qwen2.5-vl-72b-instruct with the OpenAI SDK?
Install the OpenAI SDK, set base_url to "https://router.requesty.ai/v1", set your API key to your Requesty key, and set the model to "alibaba/qwen2.5-vl-72b-instruct". The Quickstart above shows Python, JavaScript and cURL snippets.

Access qwen2.5-vl-72b-instruct through Requesty

One API key, 400+ models, OpenAI-compatible. No markup on provider prices, automatic failover, and smart caching built-in.

Get started freeAll Alibaba Cloud models