Compare · ModelsLive · 2 picked · head to head
Qwen2 0.5B vs Pythia 160m
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen2 0.5B wins on 5/6 benchmarks
Qwen2 0.5B wins 5 of 6 shared benchmarks. Leads in general · knowledge · language.
Category leads
general·Qwen2 0.5Bknowledge·Qwen2 0.5Blanguage·Qwen2 0.5Bmath·Qwen2 0.5Breasoning·Pythia 160m
Hype vs Reality
Attention vs performance
Qwen2 0.5B
#231 by perf·no signal
Pythia 160m
#235 by perf·no signal
Vendor risk
Who is behind the model
Alibaba (Qwen)
$293.0B·Tier 1
eleutherai
private · undisclosed
Head to head
6 benchmarks · 2 models
Qwen2 0.5BPythia 160m
BBH (HuggingFace)
Qwen2 0.5B leads by +5.7
Qwen2 0.5B
7.9
Pythia 160m
2.2
GPQA
Qwen2 0.5B leads by +0.3
Qwen2 0.5B
1.4
Pythia 160m
1.1
IFEval
Qwen2 0.5B leads by +0.6
Qwen2 0.5B
18.7
Pythia 160m
18.2
MATH Level 5
Qwen2 0.5B leads by +1.7
Qwen2 0.5B
2.6
Pythia 160m
0.9
MMLU-PRO
Qwen2 0.5B leads by +6.7
Qwen2 0.5B
8.0
Pythia 160m
1.3
MUSR
Pythia 160m leads by +6.1
Qwen2 0.5B
4.6
Pythia 160m
10.7
Full benchmark table
| Benchmark | Qwen2 0.5B | Pythia 160m |
|---|---|---|
BBH (HuggingFace) | 7.9 | 2.2 |
GPQA | 1.4 | 1.1 |
IFEval | 18.7 | 18.2 |
MATH Level 5 | 2.6 | 0.9 |
MMLU-PRO | 8.0 | 1.3 |
MUSR | 4.6 | 10.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| Qwen2 0.5B | — | — | — | — |
| Pythia 160m | — | — | — | — |