Google model product
Gemma 3 4B is a 4-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for question answering, summarization, reasoning, and image understanding tasks.
Updated Aug 10, 2026. Default version: Gemma 3 4B
Structured fields from the published default version.
This profile uses the latest version under this unique model that has a calculated LLMBoard score. Arena and price are excluded.
Published benchmark records for the scored version Gemma 3 4B.
| VQAv2 (val) | 0.6 | 3 | 3 | 0.0% | C | |
| ECLeKTic | 0.0 | 4 | 8 | 57.1% | C | |
| Bird-SQL (dev) | 0.4 | 6 | 7 | 16.7% | C | |
| HiddenMath | 0.4 | 7 | 13 | 50.0% | C | |
| Natural2Code | 0.7 | 7 | 8 | 14.3% | C | |
| FACTS Grounding | 0.7 | 9 | 13 | 33.3% | C | |
| InfoVQA | 0.5 | 9 | 9 | 0.0% | C | |
| BIG-Bench Extra Hard | 0.1 | 10 | 11 | 10.0% | C | |
| BIG-Bench Hard | 0.7 | 11 | 21 | 50.0% | C | |
| MMMU (val) | 0.5 | 11 | 11 | 0.0% | C | |
| Global-MMLU-Lite | 0.5 | 13 | 14 | 7.7% | C | |
| IFEval | 0.9 | 15 | 65 | 78.1% | C | |
| TextVQA | 0.6 | 15 | 15 | 0.0% | C | |
| WMT24++ | 0.5 | 18 | 23 | 22.7% | C | |
| MathVista-Mini | 0.5 | 23 | 23 | 0.0% | C | |
| ChartQA | 0.7 | 24 | 24 | 0.0% | C | |
| DocVQA | 0.8 | 26 | 26 | 0.0% | C | |
| MATH | 0.8 | 27 | 71 | 62.9% | C | |
| GSM8k | 0.9 | 28 | 48 | 42.5% | C | |
| MBPP | 0.6 | 30 | 37 | 19.4% | C | |
| AI2D | 0.7 | 31 | 32 | 3.2% | C | |
| SimpleQA | 0.0 | 43 | 46 | 6.7% | C | |
| HumanEval | 0.7 | 63 | 76 | 17.3% | C | |
| LiveCodeBench | 0.1 | 72 | 73 | 1.4% | C | |
| MMLU-Pro | 0.4 | 122 | 129 | 5.5% | C | |
| GPQA | 0.3 | 220 | 233 | 5.6% | C |
Preference and agent-evaluation signals from published Arena datasets.
| text | german | 184 | 1280.6 | 162 | N/A | |
| text | industry medicine and healthcare | 189 | 1341.6 | 197 | N/A | |
| text | industry business and management and financial operations | 199 | 1310.8 | 390 | N/A | |
| text | industry legal and government | 201 | 1327.7 | 245 | N/A | |
| text style control | industry medicine and healthcare | 203 | 1351.2 | 197 | N/A | |
| text | russian | 206 | 1294.2 | 307 | N/A | |
| text style control | german | 207 | 1289.5 | 162 | N/A | |
| text style control | industry business and management and financial operations | 208 | 1329.2 | 390 | N/A | |
| text style control | russian | 211 | 1314.5 | 307 | N/A | |
| text | industry writing and literature and language | 217 | 1288.2 | 1,080 | N/A | |
| text style control | industry legal and government | 217 | 1339.2 | 245 | N/A | |
| text | industry life and physical and social science | 219 | 1305.6 | 780 | N/A | |
| text | creative writing | 220 | 1271.0 | 666 | N/A | |
| text | non english | 220 | 1273.9 | 1,599 | N/A | |
| text | overall | 226 | 1290.9 | 4,171 | N/A | |
| text | exclude ties | 229 | 1221.4 | 2,830 | N/A | |
| text style control | industry writing and literature and language | 236 | 1297.8 | 1,080 | N/A | |
| text | english | 238 | 1300.1 | 2,572 | N/A | |
| text style control | non english | 239 | 1292.3 | 1,599 | N/A | |
| text | longer query | 240 | 1273.6 | 444 | N/A | |
| text | expert | 243 | 1222.5 | 208 | N/A | |
| text style control | creative writing | 244 | 1275.3 | 666 | N/A | |
| text style control | longer query | 246 | 1311.0 | 444 | N/A | |
| text | industry mathematical | 247 | 1255.6 | 365 | N/A | |
| text | industry entertainment and sports and media | 248 | 1233.5 | 761 | N/A | |
| text | hard prompts | 253 | 1252.8 | 1,092 | N/A | |
| text style control | industry mathematical | 253 | 1274.3 | 365 | N/A | |
| text | industry software and it services | 255 | 1269.5 | 986 | N/A | |
| text style control | overall | 255 | 1303.4 | 4,171 | N/A | |
| text | instruction following | 256 | 1239.5 | 1,233 | N/A | |
| text | multi turn | 256 | 1255.1 | 565 | N/A | |
| text style control | exclude ties | 257 | 1235.0 | 2,830 | N/A | |
| text style control | industry life and physical and social science | 257 | 1314.9 | 780 | N/A | |
| text | math | 258 | 1239.0 | 423 | N/A | |
| text | hard prompts english | 261 | 1253.5 | 764 | N/A | |
| text style control | english | 264 | 1308.6 | 2,572 | N/A | |
| text style control | expert | 265 | 1257.8 | 208 | N/A | |
| text style control | math | 265 | 1253.8 | 423 | N/A | |
| text style control | instruction following | 267 | 1267.5 | 1,233 | N/A | |
| text style control | multi turn | 267 | 1271.9 | 565 | N/A | |
| text style control | hard prompts | 269 | 1284.0 | 1,092 | N/A | |
| text | coding | 270 | 1230.8 | 605 | N/A | |
| text style control | industry entertainment and sports and media | 271 | 1248.6 | 761 | N/A | |
| text style control | industry software and it services | 273 | 1300.9 | 986 | N/A | |
| text style control | hard prompts english | 278 | 1283.8 | 764 | N/A | |
| text style control | coding | 296 | 1273.9 | 605 | N/A |
Official vendor API PAYG pricing is summarized first. The table then lists individual provider offerings without treating their minimum as the official price.
| OpenRouter | google/gemma-3-4b-it | global | $0.05 | $0.10 | 131.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
All published versions linked to this unique model. The score columns identify the version used by the current overall ranking.
| Gemma 3 4B | 22.3 | 4B | 131.1K | 131.1K | No | Gemma |
A concise description based on the published model registry.
Gemma 3 4B is a 4-billion-parameter vision-language model from Google, handling text and image input and generating text output. It features a 128K context window, multilingual support, and open weights. Suitable for question answering, summarization, reasoning, and image understanding tasks.
Use the benchmark, Arena and pricing sections above as separate evidence. A missing field means the current data snapshot does not support that claim.
Data snapshot: 2026-08-07. Editorial model content is not available in the backend.
Open a comparison with the three ranked models immediately above and below this model.
Recommendations prioritize the same model type and family, then the closest published LLMBoard score.
Common questions about Gemma 3 4B.
Gemma 3 4B's default version was released on Mar 12, 2025.
No official standard PAYG price is currently available for Gemma 3 4B. The lowest tracked third-party offer starts at $0.05 input and $0.10 output via OpenRouter.
Gemma 3 4B is published under Google in the model registry.
The default version has a 131.1K token context window.
No. The default version is not marked as having publicly available weights.
1 published provider offerings are linked to the default version.
Nearby ranked alternatives include Ministral 8B, GPT-4.1-mini, Gemma 3n E4B LiteRT.