Google model product
Gemma 3 12B is a 12-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 12B
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 12B.
| VQAv2 (val) | 0.7 | 1 | 3 | 100.0% | C | |
| ECLeKTic | 0.1 | 3 | 8 | 71.4% | C | |
| Bird-SQL (dev) | 0.5 | 4 | 7 | 50.0% | C | |
| HiddenMath | 0.5 | 4 | 13 | 75.0% | C | |
| Natural2Code | 0.8 | 4 | 8 | 57.1% | C | |
| BIG-Bench Hard | 0.9 | 6 | 21 | 75.0% | C | |
| FACTS Grounding | 0.8 | 6 | 13 | 58.3% | C | |
| Global-MMLU-Lite | 0.7 | 7 | 14 | 53.9% | C | |
| BIG-Bench Extra Hard | 0.2 | 8 | 11 | 30.0% | C | |
| InfoVQA | 0.6 | 8 | 9 | 12.5% | C | |
| MMMU (val) | 0.6 | 10 | 11 | 10.0% | C | |
| TextVQA | 0.7 | 13 | 15 | 14.3% | C | |
| MATH | 0.8 | 14 | 71 | 81.4% | C | |
| WMT24++ | 0.5 | 15 | 23 | 36.4% | C | |
| GSM8k | 0.9 | 16 | 48 | 68.1% | C | |
| DocVQA | 0.9 | 22 | 26 | 16.0% | C | |
| IFEval | 0.9 | 22 | 65 | 67.2% | C | |
| MathVista-Mini | 0.6 | 22 | 23 | 4.5% | C | |
| MBPP | 0.7 | 22 | 37 | 41.7% | C | |
| ChartQA | 0.8 | 23 | 24 | 4.3% | C | |
| AI2D | 0.8 | 24 | 32 | 25.8% | C | |
| HumanEval | 0.9 | 38 | 76 | 50.7% | C | |
| SimpleQA | 0.1 | 42 | 46 | 8.9% | C | |
| LiveCodeBench | 0.2 | 65 | 73 | 11.1% | C | |
| MMLU-Pro | 0.6 | 103 | 129 | 20.3% | C | |
| GPQA | 0.4 | 202 | 233 | 13.4% | C |
Preference and agent-evaluation signals from published Arena datasets.
| text | german | 133 | 1370.7 | 166 | N/A | |
| text style control | german | 142 | 1371.2 | 166 | N/A | |
| text | industry business and management and financial operations | 145 | 1376.3 | 364 | N/A | |
| text | industry legal and government | 149 | 1381.2 | 217 | N/A | |
| text style control | industry business and management and financial operations | 153 | 1385.7 | 364 | N/A | |
| text | creative writing | 164 | 1331.8 | 600 | N/A | |
| text style control | industry legal and government | 167 | 1382.1 | 217 | N/A | |
| text style control | russian | 169 | 1358.0 | 285 | N/A | |
| text | industry life and physical and social science | 171 | 1365.9 | 760 | N/A | |
| text | russian | 175 | 1335.8 | 285 | N/A | |
| text style control | creative writing | 176 | 1333.4 | 600 | N/A | |
| text style control | non english | 185 | 1336.6 | 1,499 | N/A | |
| text | non english | 186 | 1318.6 | 1,499 | N/A | |
| text style control | industry life and physical and social science | 189 | 1367.8 | 760 | N/A | |
| text style control | multi turn | 189 | 1344.1 | 471 | N/A | |
| text | industry mathematical | 190 | 1335.5 | 333 | N/A | |
| text | exclude ties | 192 | 1289.4 | 2,526 | N/A | |
| text | industry medicine and healthcare | 192 | 1334.1 | 155 | N/A | |
| text style control | industry mathematical | 192 | 1349.2 | 333 | N/A | |
| text | overall | 193 | 1334.3 | 3,829 | N/A | |
| text | industry entertainment and sports and media | 193 | 1298.2 | 697 | N/A | |
| text | industry writing and literature and language | 194 | 1307.3 | 1,010 | N/A | |
| text | multi turn | 194 | 1333.6 | 471 | N/A | |
| text | english | 197 | 1348.1 | 2,330 | N/A | |
| text style control | industry entertainment and sports and media | 197 | 1306.2 | 697 | N/A | |
| text style control | longer query | 197 | 1345.8 | 371 | N/A | |
| text | longer query | 201 | 1316.9 | 371 | N/A | |
| text style control | math | 204 | 1317.6 | 389 | N/A | |
| text | instruction following | 205 | 1299.7 | 1,145 | N/A | |
| text | math | 205 | 1307.2 | 389 | N/A | |
| text style control | overall | 205 | 1342.0 | 3,829 | N/A | |
| text style control | exclude ties | 205 | 1296.1 | 2,526 | N/A | |
| text style control | instruction following | 208 | 1321.3 | 1,145 | N/A | |
| text style control | industry writing and literature and language | 209 | 1313.5 | 1,010 | N/A | |
| text | hard prompts | 210 | 1309.6 | 977 | N/A | |
| text style control | english | 214 | 1348.8 | 2,330 | N/A | |
| text | hard prompts english | 215 | 1318.4 | 700 | N/A | |
| text | industry software and it services | 219 | 1313.7 | 867 | N/A | |
| text | expert | 222 | 1247.3 | 186 | N/A | |
| text style control | hard prompts | 224 | 1332.0 | 977 | N/A | |
| text style control | hard prompts english | 235 | 1337.8 | 700 | N/A | |
| text style control | industry medicine and healthcare | 237 | 1322.9 | 155 | N/A | |
| text | coding | 243 | 1281.7 | 543 | N/A | |
| text style control | industry software and it services | 249 | 1334.7 | 867 | N/A | |
| text style control | expert | 254 | 1271.7 | 186 | N/A | |
| text style control | coding | 263 | 1316.3 | 543 | 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.
| Helicone | gemma-3-12b-it | global | $0.05 | $0.10 | 131.1K | |
| NovitaAI | google/gemma-3-12b-it | global | $0.05 | $0.10 | 131.1K | |
| OpenRouter | google/gemma-3-12b-it | global | $0.05 | $0.15 | 131.1K | |
| Neon | gemma-3-12b | global | $0.15 | $0.50 | 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 12B | 39.2 | 12B | 131.1K | 131.1K | No | Gemma |
A concise description based on the published model registry.
Gemma 3 12B is a 12-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 12B.
Gemma 3 12B's default version was released on Mar 12, 2025.
No official standard PAYG price is currently available for Gemma 3 12B. The lowest tracked third-party offer starts at $0.05 input and $0.10 output via Helicone.
Gemma 3 12B 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.
4 published provider offerings are linked to the default version.
Nearby ranked alternatives include Hermes 3 70B, Claude Sonnet 3, Qwen3 30B A3B.