OpenAI-compatible — swap the base URL and it works with your existing SDK.
| Type | Price |
|---|---|
| Output | |
| Benchmark | Score | What it measures | Source | Retrieved |
|---|---|---|---|---|
| Intelligence | 4.8 | Composite score across standardised reasoning, knowledge and problem-solving evaluations, measured independently by Artificial Analysis | Source | September 22, 2026 |
| Math | 3.3 | Composite score across standardised mathematics evaluations, measured independently by Artificial Analysis | Source | September 22, 2026 |
| HellaSwag | 62.3 | Commonsense sentence completion with adversarially filtered wrong endings | Source | September 22, 2026 |
| GPQA Diamond | 19.2 | Google-proof graduate science questions (hardest subset) | Source | September 22, 2026 |
| LiveCodeBench | 1.9 | Contamination-free competitive programming problems | Source | September 22, 2026 |
| MATH | 48.0 | Competition mathematics problems across five difficulty levels | Source | September 22, 2026 |
| MMLU-Pro | 14.7 | Multi-discipline knowledge + reasoning (harder MMLU) | Source | September 22, 2026 |
| Model | Input | Output | Context | Best for |
|---|---|---|---|---|
Gemma 3 (1B) This page | ||||
| Reasoning + agents | ||||
| Balanced coding + agents | ||||
| Long-context, multimodal & agentic workflows | ||||
| Reasoning + agents |
Yes, Gemma 3 (1B) can stream responses as they are generated.
Yes, Gemma 3 (1B) supports both function calling and structured outputs.
Gemma 3 (1B) was built by Google.
Gemma 3 (1B) is accessed via the v1 endpoint.
