Skip to content
Updated minutes ago
Alibaba

Qwen 2.5 VL 72B

Alibaba · dense · 72.7B parameters · 131,072 context

Quality
50.0

Parameters

72.7B

Context Window

128K tokens

Architecture

Dense

Best GPU

H200 SXM

Cheapest API

$0.90/M

Intelligence Brief

Qwen 2.5 VL 72B is a 72.7B parameter DENSE model from Alibaba, featuring Grouped Query Attention (GQA) with 80 layers and 8,192 hidden dimensions. With a 131,072 token context window, it supports tools, vision, structured output, code, math, multilingual. The most cost-effective API deployment is via together at $0.90/M output tokens. For self-hosted inference, H200 SXM delivers optimal throughput at $2553/month.

Architecture Details

TypeDENSE
Total Parameters72.7B
Active Parameters72.7B
Layers80
Hidden Dimension8,192
Attention Heads64
KV Heads8
Head Dimension128
Vocab Size152,064

Memory Requirements

BF16 Weights

145.4 GB

FP8 Weights

72.7 GB

INT4 Weights

36.4 GB

KV-Cache per Token327680 bytes
Activation Estimate3.00 GB

GPU Compatibility Matrix

Qwen 2.5 VL 72B is compatible with 37% of GPU configurations across 41 GPUs at 3 precision levels.

BF16 (Full)
FP8 (Half)
INT4 (Quarter)
Blackwell(7 GPUs)
B200 NVL (pair)360GB
B300288GB
B100 SXM192GB
GB200 NVL72 (per GPU)192GB
Hopper(7 GPUs)
H100 NVL 94GB (per GPU pair)188GB
H200 SXM141GB
H2096GB
GH20096GB
Ada Lovelace(11 GPUs)
L40S48GB
L4048GB
RTX 6000 Ada48GB
L2048GB
Ampere(16 GPUs)
A100 80GB SXM80GB
A100 80GB PCIe80GB
A1664GB
RTX A600048GB
Legend:No fitVery tightTightModerateGoodExcellent

GPU Recommendations

H200 SXMoptimal

FP8 · 1 GPU · tensorrt-llm

100/100

score

Throughput

552.3 tok/s

Latency (ITL)

1.8ms

Est. TTFT

0ms

Cost/Month

$2553

Cost/M Tokens

$1.76

Use this config →
B200 SXMoptimal

FP8 · 1 GPU · tensorrt-llm

98/100

score

Throughput

560.0 tok/s

Latency (ITL)

1.8ms

Est. TTFT

0ms

Cost/Month

$4261

Cost/M Tokens

$2.90

Use this config →
B100 SXMoptimal

FP8 · 1 GPU · tensorrt-llm

98/100

score

Throughput

560.0 tok/s

Latency (ITL)

1.8ms

Est. TTFT

0ms

Cost/Month

$4271

Cost/M Tokens

$2.90

Use this config →

Deployment Options

API

API Deployment

together

$0.90/M

output tokens

Self-Hosted

Single GPU

H200 SXM

$2553/mo

Min VRAM: 73 GB

Scale

Multi-GPU

H100 SXM x2

560.0 tok/s

TP· $3587/mo

API Pricing Comparison

ProviderInput $/MOutput $/MBadges
together$0.90$0.90
Cheapest

Cost Analysis

ProviderInput $/MOutput $/M~Monthly Cost
togetherBest Value$0.90$0.90$9

Cost per 1,000 Requests

Short (500 tok)

$0.63

via together

Medium (2K tok)

$2.52

via together

Long (8K tok)

$9.00

via together

Performance Estimates

Throughput by GPU

H200 SXM
552.3 tok/s
B200 SXM
560.0 tok/s
B100 SXM
560.0 tok/s

VRAM Breakdown (H200 SXM, FP8)

Weights
Act
Weights 72.7 GBKV-Cache 2.7 GBActivations 24.0 GBOverhead 3.6 GB

Precision Impact

bf16

145.4 GB

weights/GPU

fp8

72.7 GB

weights/GPU

~552.3 tok/s

int4

36.4 GB

weights/GPU

Capabilities

Features

Tool Use Vision Code Math Reasoning Multilingual Structured Output

Supported Frameworks

vllmsglangtgitensorrt-llm

Supported Precisions

BF16 (default)FP8INT4

Where to Deploy Qwen 2.5 VL 72B

Similar Models

Qwen 2.5 72B

72.7B params · dense

Quality: 77

from $0.90/M

Higher qualityCompare →
Smaller contextCompare →
Smaller context, Higher qualityCompare →
Smaller context, Higher qualityCompare →

Frequently Asked Questions

How much VRAM does Qwen 2.5 VL 72B need for inference?

Qwen 2.5 VL 72B requires approximately 145.4 GB of VRAM at BF16 precision, 72.7 GB at FP8, or 36.4 GB at INT4 quantization. Additional VRAM is needed for KV-cache (327680 bytes per token) and activations (~3.00 GB).

What is the best GPU for Qwen 2.5 VL 72B?

The top recommended GPU for Qwen 2.5 VL 72B is the H200 SXM using FP8 precision. It achieves approximately 552.3 tokens/sec at an estimated cost of $2553/month ($1.76/M tokens). Score: 100/100.

How much does Qwen 2.5 VL 72B inference cost?

Qwen 2.5 VL 72B API inference starts from $0.90/M input tokens and $0.90/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.