Compare · ModelsLive · 2 picked · head to head
Qwen-1_8B vs Qwen2.5 Coder 0.5B Instruct
Side by side · benchmarks, pricing, and signals you can act on.
Winner summary
Qwen2.5 Coder 0.5B Instruct wins on 2/3 benchmarks
Qwen2.5 Coder 0.5B Instruct wins 2 of 3 shared benchmarks. Leads in math.
Category leads
knowledge·Qwen-1_8Bmath·Qwen2.5 Coder 0.5B Instruct
Hype vs Reality
Attention vs performance
Qwen-1_8B
#226 by perf·no signal
Qwen2.5 Coder 0.5B Instruct
#251 by perf·no signal
Vendor risk
Who is behind the model
Alibaba (Qwen)
$293.0B·Tier 1
Alibaba (Qwen)
$293.0B·Tier 1
Head to head
3 benchmarks · 2 models
Qwen-1_8BQwen2.5 Coder 0.5B Instruct
ARC AI2
Qwen-1_8B leads by +25.1
AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval.
Qwen-1_8B
37.6
Qwen2.5 Coder 0.5B Instruct
12.5
GSM8K
Qwen2.5 Coder 0.5B Instruct leads by +13.3
Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve.
Qwen-1_8B
21.2
Qwen2.5 Coder 0.5B Instruct
34.5
MMLU
Qwen2.5 Coder 0.5B Instruct leads by +18.4
Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge.
Qwen-1_8B
4.3
Qwen2.5 Coder 0.5B Instruct
22.7
Full benchmark table
| Benchmark | Qwen-1_8B | Qwen2.5 Coder 0.5B Instruct |
|---|---|---|
ARC AI2 AI2 Reasoning Challenge · tests grade-school level science knowledge with multiple-choice questions requiring reasoning beyond simple retrieval. | 37.6 | 12.5 |
GSM8K Grade School Math 8K · 8,500 linguistically diverse grade-school math word problems that require multi-step reasoning to solve. | 21.2 | 34.5 |
MMLU Massive Multitask Language Understanding · 57 subjects spanning STEM, humanities, social sciences, and more. The standard benchmark for broad knowledge. | 4.3 | 22.7 |
Pricing · per 1M tokens · projected $/mo at 10M tokens
| Model | Input | Output | Context | Projected $/mo |
|---|---|---|---|---|
| — | — | — | — | |
| — | — | — | — |