GGoogle · Released Mar 13, 2025
Gemma 3 4B
Gemma 3 introduces multimodality, supporting vision-language input and text outputs. It handles context windows up to 128k tokens, understands over 140 languages, and offers improved math, reasoning, and chat capabilities,...
- Released
- Mar 13 2025
- Context window
- 131K 118K out
- Input / 1M tokens
- $0.04 from Free
- Output / 1M tokens
- $0.05 from Free
Where to run it
10 providers compared
Free to try on Nvidia — free endpoints are often limited-time and rate-limited. Cheapest: Parasail, Kilo Gateway at $0.04 / $0.05 per 1M input / output tokens — 79% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| Nvidiafree · may be limited-time | Free | Free | — | 131K | 16K | — | Free |
| ParasailBF16 · no tool calling | $0.04 | $0.05 | $0.02 | 131K | 118K | 99.8%24h | Cheapest |
| Kilo Gatewayno tool calling | $0.04 | $0.05 | $0.02 | 131K | 118K | — | Cheapest |
| Amazon Bedrockno tool calling | $0.04 | $0.08 | — | 131K | 4K | — | +18% |
| Merge Gatewayno tool calling | $0.04 | $0.08 | — | 128K | 8K | — | +18% |
| Eden AI | $0.04 | $0.08 | — | 128K | 131K | — | +18% |
| DeepInfraBF16 · no tool calling | $0.05 | $0.10 | — | 131K | 16K | 100%24h | +47% |
| Hugging Face | $0.05 | $0.10 | — | 131K | 131K | — | +47% |
| AI/ML API | $0.069 | $0.14 | $0.055 | 131K | 33K | — | +103% |
| NanoGPTno tool calling | $0.20 | $0.20 | $0.10 | 128K | 8K | — | +373% |
Value
Price vs. performance
Gemma 3 4B Other Google Other labs Value frontier · GPQA Diamond vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
3 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| GPQA DiamondReasoning & knowledge · PhD-level biology, chemistry and physics questions | 23.2% | #122of 122 | Gemini 4 Argon96.4% |
| LMCAReasoning & knowledge · Judge the quality of conceptual arguments like experts | 2.8% | #123of 123 | Claude Sonnet 5.568.3% |
| OTIS Mock AIMEMaths · Competition-style problems harder than MATH Level 5 | 7.5% | #104of 120 | Gemini 4 Argon100.0% |
Specs
Details
- Model ID
google/gemma-3-4b-it- Developer
- Release date
- Mar 13, 2025 ·
- Context window
- 131,072 tokens
- Max output
- 117,964 tokens
- API price
- $0.04 / $0.05 per 1M input / output tokens
- Providers
- 10
- Input types
- text, image
- Reasoning
- No
- Tool calling
- No
- Open weights
- Yes
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.