Compare / Gemma 3 270M vs Gemma 3 1B
Gemma 3 270M vs Gemma 3 1B
Updated November 2026
Verdict
Gemma 3 270M scores higher on the Small LM Index: 12.8 vs 8.9. On the 8 benchmarks both labs report, Gemma 3 270M wins 1 and Gemma 3 1B wins 6. Both need about 1.2 GB of memory at 4-bit quantization. Neither model has a hosted API on OpenRouter, so you would run both yourself.
Pick Gemma 3 270M if…
- you want the strongest model overall
- you need math and reasoning
Pick Gemma 3 1B if…
- you mostly write code
- you need precise instruction following
- you build agents that call tools
- you want the more widely used model (more quantized builds and community support)
Specifications
| Gemma 3 270M | Gemma 3 1B | |
|---|---|---|
| Small LM Index | 12.8 (likely 12–14) | 8.9 (likely 8–10) |
| Lab | Google · United States | Google · United States |
| Released | 2025-08-14 | 2025-03-12 |
| Parameters | 270M (dense) | 1B (dense) |
| Memory at 4-bit | ~0.7 GB | ~1.2 GB |
| Context window | 32K tokens | 32K tokens |
| Input | text | text |
| Reasoning | No thinking | No thinking |
| License | Gemma | Gemma |
| API price / 1M tokens | Self-host only | Self-host only |
| Hugging Face downloads (30 days) | 60.4K | 188.4K |
Benchmarks
Higher is better. Bold marks the winner where both labs report a score. ≈ marks the index model's estimate for a score that wasn't reported, ✓ an independent measurement and * a score measured by another lab.
| Benchmark | Gemma 3 270M | Gemma 3 1B |
|---|---|---|
| MMLU-Pro Knowledge | 5.5✓ | 13.5✓ |
| GPQA Diamond Knowledge | 22.4✓ | 23.7✓ |
| Humanity's Last Exam Reasoning | 3.6✓ | 5.3✓ |
| AIME Math | 2.3✓ | 0.0✓ |
| LiveCodeBench Coding | 0.3✓ | 1.7✓ |
| τ²-Bench Agentic | 9.1✓ | 10.5✓ |
| IFBench Instruction following | 12.1✓ | 19.9✓ |
| MATH-500 Math | — | 48.4✓ |
| AA-LCR Long context | 0.0✓ | 0.0✓ |
Scores are mostly reported by each lab with its own settings, so small gaps are noise. See the methodology. Open in the interactive comparison to add more models.
Questions
Which is better, Gemma 3 270M or Gemma 3 1B?
Gemma 3 270M scores higher on the Small LM Index: 12.8 vs 8.9. On the 8 benchmarks both labs report, Gemma 3 270M wins 1 and Gemma 3 1B wins 6.
Which is cheaper to run?
Neither model has a hosted API on OpenRouter, so you would run both yourself. Both need about 1.2 GB of memory at 4-bit quantization.
Which has the longer context window?
Both have a 32K-token context window.
Can I use Gemma 3 270M and Gemma 3 1B commercially?
Both are released under the Gemma license. Apache 2.0 and MIT allow commercial use; other licenses can add conditions, so read the license before shipping a product.