Compare · ModelsLive · 2 picked · head to head
Qwen2.5 72B Instruct vs DeepSeek R1 Distill Qwen 14B
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen2.5 72B Instruct wins on 4/6 benchmarks
Qwen2.5 72B Instruct wins 4 of 6 shared benchmarks. Leads in general · language · math.
Category leads
general·Qwen2.5 72B Instructknowledge·DeepSeek R1 Distill Qwen 14Blanguage·Qwen2.5 72B Instructmath·Qwen2.5 72B Instructreasoning·DeepSeek R1 Distill Qwen 14B
Hype vs Reality
Attention vs performance
Qwen2.5 72B Instruct
#80 by perf·no signal
DeepSeek R1 Distill Qwen 14B
#62 by perf·no signal
Best value
Qwen2.5 72B Instruct
Qwen2.5 72B Instruct
208.6 pts/$
$0.26/M
DeepSeek R1 Distill Qwen 14B
—
no price
Vendor risk
Mixed exposure
One or more vendors flagged
Alibaba (Qwen)
$293.0B·Tier 1
DeepSeek
$3.4B·Tier 1
Head to head
6 benchmarks · 2 models
Qwen2.5 72B InstructDeepSeek R1 Distill Qwen 14B
BBH (HuggingFace)
Qwen2.5 72B Instruct leads by +21.2
Qwen2.5 72B Instruct
61.9
DeepSeek R1 Distill Qwen 14B
40.7
GPQA
DeepSeek R1 Distill Qwen 14B leads by +1.7
Qwen2.5 72B Instruct
16.7
DeepSeek R1 Distill Qwen 14B
18.3
IFEval
Qwen2.5 72B Instruct leads by +42.6
Qwen2.5 72B Instruct
86.4
DeepSeek R1 Distill Qwen 14B
43.8
MATH Level 5
Qwen2.5 72B Instruct leads by +2.8
Qwen2.5 72B Instruct
59.8
DeepSeek R1 Distill Qwen 14B
57.0
MMLU-PRO
Qwen2.5 72B Instruct leads by +10.7
Qwen2.5 72B Instruct
51.4
DeepSeek R1 Distill Qwen 14B
40.7
MUSR
DeepSeek R1 Distill Qwen 14B leads by +17.0
Qwen2.5 72B Instruct
11.7
DeepSeek R1 Distill Qwen 14B
28.7
Full benchmark table
| Benchmark | Qwen2.5 72B Instruct | DeepSeek R1 Distill Qwen 14B |
|---|---|---|
BBH (HuggingFace) | 61.9 | 40.7 |
GPQA | 16.7 | 18.3 |
IFEval | 86.4 | 43.8 |
MATH Level 5 | 59.8 | 57.0 |
MMLU-PRO | 51.4 | 40.7 |
MUSR | 11.7 | 28.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| $0.12 | $0.39 | 33K tokens (~16 books) | $1.88 | |
| — | — | — | — |