Qwen 2.5 72B vs DeepSeek R1
Architecture Comparison
SpecQwen 2.5 72BDeepSeek R1
TypeDENSEMOE
Total Parameters72.7B671B
Active Parameters72.7B37B
Layers8061
Hidden Dimension8,1927,168
Attention Heads64128
KV Heads81
Context Length131,072131,072
Precision (default)BF16BF16
Total ExpertsN/A256
Active ExpertsN/A8
Memory Requirements
PrecisionQwen 2.5 72BDeepSeek R1
BF16 Weights145.4 GB1342.0 GB
FP8 Weights72.7 GB671.0 GB
INT4 Weights36.4 GB335.5 GB
KV-Cache / Token327680 B31232 B
Activation Estimate2.50 GB3.00 GB
Minimum GPUs Needed (BF16)
H100 SXM3 GPUsN/A
L40S4 GPUsN/A
Quality Benchmarks
BenchmarkQwen 2.5 72BDeepSeek R1
Overall8492
MMLU85.390.8
HumanEval56.071.7
GSM8K91.697.3
MT-Bench86.089.0
Qwen 2.5 72B
MMLU
85.3
HumanEval
56.0
GSM8K
91.6
MT-Bench
86.0
DeepSeek R1
MMLU
90.8
HumanEval
71.7
GSM8K
97.3
MT-Bench
89.0
Capabilities
FeatureQwen 2.5 72BDeepSeek R1
Tool Use✓ Yes✓ Yes
Vision✗ No✗ No
Code✓ Yes✓ Yes
Math✓ Yes✓ Yes
Reasoning✗ No✓ Yes
Multilingual✓ Yes✓ Yes
Structured Output✓ Yes✓ Yes
API Pricing Comparison
Cheapest Output (Qwen 2.5 72B)
$0.90/M
Input: $0.90/M
Cheapest Output (DeepSeek R1)
$2.19/M
Input: $0.55/M
| Provider | Qwen 2.5 72B In $/M | Out $/M | DeepSeek R1 In $/M | Out $/M |
|---|---|---|---|---|
| together | $0.90 | $0.90 | $3.00 | $7.00 |
| fireworks | $0.90 | $0.90 | — | — |
| deepseek | — | — | $0.55 | $2.19 |
Recommendation Summary
- ‣DeepSeek R1 scores higher on overall quality (92 vs 84).
- ‣Qwen 2.5 72B is cheaper per output token ($0.90/M vs $2.19/M).
- ‣Qwen 2.5 72B has a smaller memory footprint (145.4 GB vs 1342.0 GB BF16), making it easier to deploy on fewer GPUs.
- ‣Qwen 2.5 72B uses DENSE architecture while DeepSeek R1 uses MOE. MoE models activate fewer parameters per token, improving inference efficiency.
- ‣DeepSeek R1 is stronger at code generation (HumanEval: 71.7 vs 56.0).
- ‣DeepSeek R1 is better at math reasoning (GSM8K: 97.3 vs 91.6).