Qwen 2.5 Coder 32B
Alibaba · dense · 32.5B parameters · 131,072 context
Parameters
32.5B
Context Window
128K tokens
Architecture
Dense
Best GPU
H20
Cheapest API
$0.80/M
Quality Score
80/100
Intelligence Brief
Qwen 2.5 Coder 32B is a 32.5B parameter DENSE model from Alibaba, featuring Grouped Query Attention (GQA) with 64 layers and 5,120 hidden dimensions. With a 131,072 token context window, it supports tools, structured output, code, math, multilingual. On standardized benchmarks, it achieves MMLU 74, HumanEval 92.7, GSM8K 76. The most cost-effective API deployment is via alibaba at $0.80/M output tokens. For self-hosted inference, H20 delivers optimal throughput at $940/month.
Architecture Details
Memory Requirements
BF16 Weights
65.0 GB
FP8 Weights
32.5 GB
INT4 Weights
16.3 GB
GPU Compatibility Matrix
Qwen 2.5 Coder 32B is compatible with 57% of GPU configurations across 41 GPUs at 3 precision levels.
GPU Recommendations
FP8 · 1 GPU · tensorrt-llm
100/100
score
Throughput
1.0K tok/s
Latency (ITL)
1.0ms
Est. TTFT
0ms
Cost/Month
$940
Cost/M Tokens
$0.35
FP8 · 1 GPU · tensorrt-llm
95/100
score
Throughput
1.1K tok/s
Latency (ITL)
1.0ms
Est. TTFT
0ms
Cost/Month
$2553
Cost/M Tokens
$0.93
FP8 · 1 GPU · tensorrt-llm
95/100
score
Throughput
862.3 tok/s
Latency (ITL)
1.2ms
Est. TTFT
0ms
Cost/Month
$1794
Cost/M Tokens
$0.79
Deployment Options
API Deployment
alibaba
$0.80/M
output tokens
Single GPU
H20
$940/mo
Min VRAM: 33 GB
Multi-GPU
RTX A6000 x2
112.2 tok/s
TP· $930/mo
API Pricing Comparison
| Provider | Input $/M | Output $/M | Badges |
|---|---|---|---|
| alibaba | $0.80 | $0.80 | Cheapest |
| together | $0.80 | $0.80 |
Cost Analysis
| Provider | Input $/M | Output $/M | ~Monthly Cost |
|---|---|---|---|
| alibabaBest Value | $0.80 | $0.80 | $8 |
| together | $0.80 | $0.80 | $8 |
Cost per 1,000 Requests
Short (500 tok)
$0.56
via alibaba
Medium (2K tok)
$2.24
via alibaba
Long (8K tok)
$8.00
via alibaba
Performance Estimates
Throughput by GPU
VRAM Breakdown (H20, FP8)
Precision Impact
bf16
65.0 GB
weights/GPU
fp8
32.5 GB
weights/GPU
~1.0K tok/s
int4
16.3 GB
weights/GPU
Quality Benchmarks
Capabilities
Features
Supported Frameworks
Supported Precisions
Where to Deploy Qwen 2.5 Coder 32B
Self-Hosted Infrastructure
Similar Models
Qwen 2.5 32B
32.5B params · dense
Quality: 73
from $0.80/M
Qwen 2.5 14B
14.8B params · dense
Quality: 76
from $0.30/M
Qwen 2.5 72B
72.7B params · dense
Quality: 77
from $0.90/M
Qwen 2.5 Coder 32B
32.5B params · dense
Quality: 50
from $0.80/M
DeepSeek R1 Distill 32B
32.8B params · dense
Quality: 88
from $0.50/M
Frequently Asked Questions
How much VRAM does Qwen 2.5 Coder 32B need for inference?
Qwen 2.5 Coder 32B requires approximately 65.0 GB of VRAM at BF16 precision, 32.5 GB at FP8, or 16.3 GB at INT4 quantization. Additional VRAM is needed for KV-cache (262144 bytes per token) and activations (~2.00 GB).
What is the best GPU for Qwen 2.5 Coder 32B?
The top recommended GPU for Qwen 2.5 Coder 32B is the H20 using FP8 precision. It achieves approximately 1.0K tokens/sec at an estimated cost of $940/month ($0.35/M tokens). Score: 100/100.
How much does Qwen 2.5 Coder 32B inference cost?
Qwen 2.5 Coder 32B API inference starts from $0.80/M input tokens and $0.80/M output tokens. Self-hosted inference costs depend on your GPU configuration — use our ROI calculator for a detailed breakdown.