Compare · ModelsLive · 2 picked · head to head
DeepSeek R1 Distill Qwen 32B vs Meta Llama 3 8B Instruct
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
DeepSeek R1 Distill Qwen 32B wins on 8/11 benchmarks
DeepSeek R1 Distill Qwen 32B wins 8 of 11 shared benchmarks. Leads in knowledge · language · math.
Category leads
general·Meta Llama 3 8B Instructknowledge·DeepSeek R1 Distill Qwen 32Blanguage·DeepSeek R1 Distill Qwen 32Bmath·DeepSeek R1 Distill Qwen 32Breasoning·DeepSeek R1 Distill Qwen 32B
Head to head · from the data · data as of 2026-11-22
- Benchmarks · DeepSeek R1 Distill Qwen 32B leads on 8 of 11 shared benchmarks (GPQA, MATH Level 5, MMLU-PRO, MUSR, 4 more), Meta Llama 3 8B Instruct on 3 (BBH (HuggingFace), IFEval, JSQuAD)
- Released · DeepSeek R1 Distill Qwen 32B Jan 20, 2025 vs Meta Llama 3 8B Instruct Apr 17, 2024 · 9 months apart
- Open weights · DeepSeek R1 Distill Qwen 32B and Meta Llama 3 8B Instruct both have open weights
Sources · scores from public leaderboards, list prices and provider offers from OpenRouter, price moves from the BenchGecko changelog (5% or more, held in the next daily check), Gecko tests run by BenchGecko.
Hype vs Reality
Attention vs performance
DeepSeek R1 Distill Qwen 32B
#177 by perf·no signal
Meta Llama 3 8B Instruct
#176 by perf·no signal
Best value
Pricing unknown
DeepSeek R1 Distill Qwen 32B
n/a
no price
Meta Llama 3 8B Instruct
n/a
no price
Vendor risk
Mixed exposure
One or more vendors flagged
DeepSeek
$3.4B·Tier 1
Meta AI
$1.44T·Tier 1
Head to head
11 benchmarks · 2 models
DeepSeek R1 Distill Qwen 32BMeta Llama 3 8B Instruct
BBH (HuggingFace)
Meta Llama 3 8B Instruct leads by +11.1
DeepSeek R1 Distill Qwen 32B
17.1
Meta Llama 3 8B Instruct
28.2
GPQA
DeepSeek R1 Distill Qwen 32B leads by +3.4
DeepSeek R1 Distill Qwen 32B
4.6
Meta Llama 3 8B Instruct
1.2
IFEval
Meta Llama 3 8B Instruct leads by +32.2
DeepSeek R1 Distill Qwen 32B
41.9
Meta Llama 3 8B Instruct
74.1
MATH Level 5
DeepSeek R1 Distill Qwen 32B leads by +8.4
DeepSeek R1 Distill Qwen 32B
17.1
Meta Llama 3 8B Instruct
8.7
MMLU-PRO
DeepSeek R1 Distill Qwen 32B leads by +11.4
DeepSeek R1 Distill Qwen 32B
41.0
Meta Llama 3 8B Instruct
29.6
MUSR
DeepSeek R1 Distill Qwen 32B leads by +14.5
DeepSeek R1 Distill Qwen 32B
16.1
Meta Llama 3 8B Instruct
1.6
JCommonsenseQA
DeepSeek R1 Distill Qwen 32B leads by +7.6
DeepSeek R1 Distill Qwen 32B
95.3
Meta Llama 3 8B Instruct
87.7
JMMLU
DeepSeek R1 Distill Qwen 32B leads by +25.5
DeepSeek R1 Distill Qwen 32B
72.2
Meta Llama 3 8B Instruct
46.7
JNLI
DeepSeek R1 Distill Qwen 32B leads by +24.5
DeepSeek R1 Distill Qwen 32B
85.6
Meta Llama 3 8B Instruct
61.1
JSQuAD
Meta Llama 3 8B Instruct leads by +0.3
DeepSeek R1 Distill Qwen 32B
89.2
Meta Llama 3 8B Instruct
89.5
LLM-JP · Overall
DeepSeek R1 Distill Qwen 32B leads by +9.7
DeepSeek R1 Distill Qwen 32B
59.3
Meta Llama 3 8B Instruct
49.6
Full benchmark table
| Benchmark | DeepSeek R1 Distill Qwen 32B | Meta Llama 3 8B Instruct |
|---|---|---|
BBH (HuggingFace) | 17.1 | 28.2 |
GPQA | 4.6 | 1.2 |
IFEval | 41.9 | 74.1 |
MATH Level 5 | 17.1 | 8.7 |
MMLU-PRO | 41.0 | 29.6 |
MUSR | 16.1 | 1.6 |
JCommonsenseQA | 95.3 | 87.7 |
JMMLU | 72.2 | 46.7 |
JNLI | 85.6 | 61.1 |
JSQuAD | 89.2 | 89.5 |
LLM-JP · Overall | 59.3 | 49.6 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| n/a | n/a | n/a | n/a | |
| n/a | n/a | n/a | n/a |
DeepSeek R1 Distill Qwen 32B vs Meta Llama 3 8B Instruct · questions answered from the data
DeepSeek R1 Distill Qwen 32B leads on 8 of 11 shared benchmarks (GPQA, MATH Level 5, MMLU-PRO, MUSR, 4 more), Meta Llama 3 8B Instruct on 3 (BBH (HuggingFace), IFEval, JSQuAD)