What is Qwen2.5-VL-72B Instruct good for?

Reading documents, charts and diagrams from images and video, and visual agent tasks such as computer and phone use. Open weights under the Qwen License, on a single high-memory GPU.

When is Qwen2.5-VL-72B Instruct not a good fit?

Audio input, and long-document work, which its short context rules out. The Qwen License rather than Apache 2.0 needs checking before commercial use.

What is the cheapest way to run Qwen2.5-VL-72B Instruct?

Hosted, unless you push serious volume. Lyceum charges $0.25 in / $0.75 out per 1M tokens. The cheapest rental that fits is 4x RTX 4060 Ti at $317 a month, which costs the same as about 950M tokens a month on that API.

Can I self-host Qwen2.5-VL-72B Instruct?

Yes. The weights are released under the Qwen License. At 4-bit it needs about 48 GB of GPU memory, which starts at roughly $317 a month on the cheapest rental that fits.

Open weights · Qwen License

Qwen2.5-VL-72B Instruct

The flagship 72B Qwen2.5-VL vision-language model, notably good at document, chart and diagram understanding.

Jump to: Specs Pricing Self-hosting Similar models FAQ More from Alibaba Cloud

Visit website Compare vs other LLMs →

Cheapest via Lyceum

Input: $0.25 /1M tokens
Output: $0.75 /1M tokens

2 providers

Key Specifications

Context window: 33K tokens
Max output: 8K tokens
Released: Jan 2025
Parameters: 73.4B
Inputs: Text, image, video

Capabilities Show details: Function calling Connect to external tools, APIs, and systems. Structured output Return responses in structured formats like JSON.

Hosted API pricing

Input tokens M Output tokens M

Provider	Input /1M tokens	Output /1M tokens	Cost at 10M in + 2M out
Lyceum	$0.25	$0.75	$4.00	View
Alibaba Cloud Creator	Open weights, no hosted price			View

Heads up: Base-tier, on-demand rates per 1M tokens; cached, batch and long-context tiers excluded. A provider may serve a shorter context or a quantized build than the creator's release. Verify before provisioning. More on how we price.

Estimated cost to self-host

Qwen2.5-VL-72B Instruct needs about 48 GB of GPU memory at 4-bit with its full 32K context. The cheapest rental that fits is 4× RTX 4060 Ti at about $317 a month.

That costs the same as roughly 950M tokens a month on Lyceum's API. Self-hosting is more expensive below that volume.

Single node ≤ 8

Precision	Memory	Cheapest fit (Full 32K context)	Cost /mo	Break-even vs API
4-bitINT4 / FP4	48 GB	4× RTX 4060 Ti	$317	950M tokens /mo
8-bitFP8 / INT8	86 GB	8× RTX 3060	$346	1B tokens /mo
16-bitFP16 / BF16	160 GB	8× RTX 3090	$979	3B tokens /mo

Estimates based on median on-demand rates for Nvidia GPUs. Memory is weights plus KV cache for one request, using FP8 KV cache where supported and FP16/BF16 otherwise. Break-even assumes a 5:1 input-to-output ratio. No guarantee of runtime support, usable performance, or that a matching quantized build exists. How we estimate costs.

Similarly priced models

The models nearest Qwen2.5-VL-72B Instruct by blended rate, each at its own cheapest provider.

Model	Blended /1M	Input /1M	Output /1M	Context	Cutoff	vs Qwen2.5-VL-72B Instruct
Qwen3.5-397B-A17B Alibaba Cloud	$0.3133	$0.17	$1.03	262K		−6%
MiniMax-M3 MiniMax	$0.3167	$0.20	$0.90	1M		−5%
Qwen3-Next-80B-A3B Instruct Alibaba Cloud	$0.325	$0.15	$1.20	262K		−2%
Qwen3-Next-80B-A3B Thinking Alibaba Cloud	$0.325	$0.15	$1.20	262K		−2%
Qwen-MT-Plus Alibaba Cloud	$0.3333	$0.25	$0.75	16K		0%
Qwen2.5-VL-72B Instruct This model Alibaba Cloud	$0.3333	$0.25	$0.75	33K
DeepSeek V3.2 DeepSeek	$0.3417	$0.25	$0.80	131K		+3%
Qwen3.6-27B Alibaba Cloud	$0.35	$0.30	$0.60	262K		+5%
GPT-5.6 Luna OpenAI	$0.3667	$0.20	$1.20	1M	Feb 2026	+10%
GPT-5.4 Nano OpenAI	$0.375	$0.20	$1.25	400K	Aug 2025	+13%
Qwen2.5-72B Instruct Alibaba Cloud	$0.3833	$0.38	$0.40	33K		+15%

Prices are USD per 1M tokens at each model's cheapest listed provider. Blended is the cost of 10M input plus 2M output tokens, spread over the 12M.

Runpod

Spin Up a GPU in Seconds - No Setup Required

Deploy AI Models Globally in 30+ Regions

Scale from 0 to 1000 GPUs on Demand

PyTorch, Tensorflow & Docker Ready

Learn more

Frequently Asked Questions

What is Qwen2.5-VL-72B Instruct good for?: Reading documents, charts and diagrams from images and video, and visual agent tasks such as computer and phone use. Open weights under the Qwen License, on a single high-memory GPU.
When is Qwen2.5-VL-72B Instruct not a good fit?: Audio input, and long-document work, which its short context rules out. The Qwen License rather than Apache 2.0 needs checking before commercial use.
What is the cheapest way to run Qwen2.5-VL-72B Instruct?: Hosted, unless you push serious volume. Lyceum charges $0.25 in / $0.75 out per 1M tokens. The cheapest rental that fits is 4x RTX 4060 Ti at $317 a month, which costs the same as about 950M tokens a month on that API.
Can I self-host Qwen2.5-VL-72B Instruct?: Yes. The weights are released under the Qwen License. At 4-bit it needs about 48 GB of GPU memory, which starts at roughly $317 a month on the cheapest rental that fits.

More from Alibaba Cloud

Model	Context	Input /1M	Output /1M
Qwen-MT-Plus	16K	$0.25	$0.75
Qwen3-Next-80B-A3B Instruct Open weights	262K	$0.15	$1.20
Qwen3-Next-80B-A3B Thinking Open weights	262K	$0.15	$1.20
Qwen3.6-27B Open weights	262K	$0.30	$0.60
Qwen3.5-397B-A17B Open weights	262K	$0.17	$1.03
Qwen2.5-7B Instruct Open weights	33K	$0.30	$0.30
Qwen3-235B-A22B FP8 Open weights	33K	$0.20	$0.80
Qwen2.5-72B Instruct Open weights	33K	$0.38	$0.40