impulse

Qwen2.5 VL 72B Instruct

Open weights

Qwen2.5-VL is proficient in recognizing common objects such as flowers, birds, fish, and insects. It is also highly capable of analyzing texts, charts, icons, graphics, and layouts within images.

API id Qwen2.5-VL-72B-InstructReleased 1 Feb 2025 (Feb 2025)
Input / 1M
$0.25
at Nebius AI Studio
Output / 1M
$0.75
at Nebius AI Studio
Context window
32K
tokens
Max output
32K
tokens per reply
Hosts
5
5 listings
Released
Feb 2025
1 Feb 2025

Pricing

List prices at Nebius AI Studio, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.25$0.75

What it costs a month

At Nebius AI Studio list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$131
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$147
  • Coding agent1,000 requests a day, 40K in / 2K out$345
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$780
Use your own numbers

Where to run it

5 hosts serve Qwen2.5 VL 72B Instruct. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Qwen2.5-VL-72B-Instruct
$0.25 / $0.75
Novita AIInference
qwen2.5-vl-72b-instruct
$0.80 / $0.80
+113% vs headline
Fireworks AIInference
qwen2p5-vl-72b-instruct
$0.90 / $0.90
+140% vs headline
Qwen2.5-VL-72B-Instruct
$0.91 / $0.91
+143% vs headline
Together AIInference
Qwen2.5-VL-72B-Instruct
$1.95 / $8.00
+823% vs headline

Capabilities

Core

  • Streaming
  • Function calling
  • JSON mode
  • Structured outputs

Multimodal

  • Vision
  • Image input
  • Audio
  • Image generation
  • Video

Advanced

  • Reasoning
  • Tool calling
  • MCP
  • Prompt caching
  • Batch API
  • Embeddings
Takes in
TextImages
Produces
Text
Open weights
License: Apache-2.0
Runs in about 176.2 GB of GPU memory
Runtimes: Ollama, SGLang, llama.cpp, vLLM

Cheaper alternatives

Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.

Same model, cheaper host

No host sells it for less than Nebius AI Studio.

Smaller Alibaba Qwen models

No cheaper Alibaba Qwen model keeps its key capabilities.

Other frontier labs

  • Claude Haiku 5.5
    Anthropic · on Anthropic
    $0.10 / $0.50
    Has every key capability
    47% less
  • GPT-6 Luna
    OpenAI · on OpenAI
    $0.10 / $0.50
    Has every key capability
    47% less
  • GPT-4o Mini
    OpenAI · on OpenAI
    $0.15 / $0.60
    Has every key capability
    30% less

Open-weight picks

  • Qwen3.5 397B A17B
    Alibaba Qwen · on AIHubMix
    $0.164 / $0.986
    Has every key capability
    1% less
  • Muse Glimmer 30B
    Meta · on EmpirioLabs AI
    $0.20 / $0.80
    Has every key capability
    7% less
  • Mistral Small 2603
    Mistral AI · on Mistral AI
    $0.15 / $0.60
    Has every key capability
    30% less
  • Deepseekv4.1 Flash
    DeepSeek · on Sarvam
    $0.149 / $0.585
    Has every key capability
    31% less

Model line

Newer and older versions, at each lab's own price.

  1. Feb 2025Qwen2.5 VL 72B InstructLatest$0.25 / $0.75
  2. UndatedQwen2 VL 72B Instruct$0.13 / $0.40

Call it

Using Nebius AI Studio's API.

Python
from litellm import completion

response = completion(
    model="nebius/Qwen2.5-VL-72B-Instruct",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Qwen2.5 VL 72B Instruct cost?

On Nebius AI Studio, Qwen2.5 VL 72B Instruct costs $0.25 per million input tokens and $0.75 per million output tokens.

Which APIs offer Qwen2.5 VL 72B Instruct?

Qwen2.5 VL 72B Instruct runs on 5 APIs: Nebius AI Studio, Novita AI, Fireworks AI, OVHcloud AI Endpoints and Together AI.

What is Qwen2.5 VL 72B Instruct's context window?

32K tokens, with up to 32K tokens in one reply.

Is Qwen2.5 VL 72B Instruct being retired?

No retirement date is scheduled for it on any API we track.

Is there a newer version of Qwen2.5 VL 72B Instruct?

No. Qwen2.5 VL 72B Instruct is the newest model in its line.

What are cheaper alternatives to Qwen2.5 VL 72B Instruct?

Models that cost less and keep most of what it can do: Claude Haiku 5.5 on Anthropic (47% less), GPT-6 Luna on OpenAI (47% less) and Qwen3.5 397B A17B on AIHubMix (1% less). Savings are on a mix of 3 input tokens to 1 output token.

When was Qwen2.5 VL 72B Instruct released?

Alibaba Qwen released it on 1 Feb 2025.

What is Qwen2.5 VL 72B Instruct's API model id?

Use Qwen2.5-VL-72B-Instruct with Nebius AI Studio. Other hosts use their own ids: qwen2.5-vl-72b-instruct on Novita AI, qwen2p5-vl-72b-instruct on Fireworks AI.

Sources

Raw catalog entry for Qwen2.5-VL-72B-Instruct on Nebius AI Studio
{
  "max_input_tokens": 32000,
  "max_output_tokens": 32000,
  "input_cost_per_token": 2.5e-7,
  "output_cost_per_token": 7.5e-7,
  "litellm_provider": "nebius",
  "mode": "chat",
  "supports_function_calling": true,
  "supports_vision": true,
  "source": "https://tokenfactory.nebius.com/models/catalog/image2text/Qwen%2FQwen2.5-VL-72B-Instruct"
}