Google model product
Gemini 3 Pro is the first model in the new Gemini 3 series. It is best for complex tasks that require broad world knowledge and advanced reasoning across modalities. Gemini 3 Pro uses dynamic thinking by default to reason through prompts, and features a 1 million-token input context window with 64k output tokens.
Updated Aug 10, 2026. Default version: Gemini 3 Pro
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 Gemini 3 Pro.
| AIME 2025 | 1.0 | 1 | 114 | 100.0% | C | |
| Global PIQA | 0.9 | 1 | 13 | 100.0% | C | |
| VideoMMMU | 0.9 | 1 | 26 | 100.0% | C | |
| LiveCodeBench Pro | 2439.0 | 2 | 4 | 66.7% | C | |
| Vending-Bench 2 | 5478.2 | 3 | 4 | 33.3% | C | |
| MMMLU | 0.9 | 4 | 49 | 93.8% | C | |
| ScreenSpot Pro | 0.7 | 6 | 24 | 78.3% | C | |
| SimpleQA | 0.7 | 6 | 46 | 88.9% | C | |
| MathArena Apex | 0.2 | 7 | 7 | 0.0% | C | |
| t2-bench | 0.9 | 7 | 23 | 72.7% | C | |
| FACTS Grounding | 0.7 | 8 | 13 | 41.7% | C | |
| MMMU-Pro | 0.8 | 10 | 65 | 85.9% | C | |
| ARC-AGI v2 | 0.3 | 12 | 16 | 26.7% | C | |
| MRCR v2 (8-needle) | 0.3 | 14 | 21 | 35.0% | C | |
| GPQA | 0.9 | 16 | 233 | 93.5% | C | |
| OmniDocBench 1.5 | 0.1 | 16 | 16 | 0.0% | C | |
| CharXiv-R | 0.8 | 17 | 47 | 65.2% | C | |
| LiveBench | 0.7 | 21 | 38 | 46.0% | B | |
| Humanity's Last Exam | 0.5 | 26 | 92 | 72.5% | C | |
| Terminal-Bench 2.0 | 0.5 | 30 | 49 | 39.6% | C | |
| SWE-Bench Verified | 0.8 | 33 | 104 | 68.9% | C |
Preference and agent-evaluation signals from published Arena datasets.
| text | german | 1 | 1515.9 | 762 | N/A | |
| text | japanese | 1 | 1512.2 | 409 | N/A | |
| text factuality | german | 1 | 1475.7 | 347 | N/A | |
| text factuality | polish | 1 | 1487.3 | 850 | N/A | |
| text style control | german | 1 | 1520.8 | 762 | N/A | |
| vision | creative writing | 1 | 1390.6 | 766 | N/A | |
| vision | entity recognition | 1 | 1354.1 | 240 | N/A | |
| vision style control | creative writing | 1 | 1372.9 | 766 | N/A | |
| vision style control | creative writing vision | 1 | 1331.3 | 1,031 | N/A | |
| vision style control | entity recognition | 1 | 1299.0 | 240 | N/A | |
| text factuality | industry medicine and healthcare | 2 | 1522.3 | 1,451 | N/A | |
| text factuality | spanish | 2 | 1488.6 | 672 | N/A | |
| text style control | polish | 2 | 1504.6 | 1,291 | N/A | |
| vision | captioning | 2 | 1318.3 | 216 | N/A | |
| vision | creative writing vision | 2 | 1348.3 | 1,031 | N/A | |
| vision style control | captioning | 2 | 1271.2 | 216 | N/A | |
| text | polish | 3 | 1502.3 | 1,291 | N/A | |
| text style control | japanese | 3 | 1509.2 | 409 | N/A | |
| vision | humor | 3 | 1338.2 | 564 | N/A | |
| vision style control | humor | 3 | 1316.1 | 564 | N/A | |
| text factuality | creative writing | 5 | 1477.7 | 3,775 | N/A | |
| text style control | creative writing | 5 | 1484.4 | 6,278 | N/A | |
| text style control | russian | 5 | 1502.4 | 4,036 | N/A | |
| text style control | industry medicine and healthcare | 6 | 1508.9 | 2,532 | N/A | |
| vision | chinese | 6 | 1375.9 | 1,388 | N/A | |
| text factuality | russian | 7 | 1495.8 | 2,237 | N/A | |
| text | industry medicine and healthcare | 8 | 1492.4 | 2,532 | N/A | |
| text | russian | 8 | 1495.1 | 4,036 | N/A | |
| text factuality | exclude ties | 8 | 1493.4 | 28,867 | N/A | |
| text factuality | industry writing and literature and language | 8 | 1475.2 | 5,951 | N/A | |
| text factuality | math | 8 | 1481.0 | 1,503 | N/A | |
| text factuality | non english | 8 | 1474.1 | 21,598 | N/A | |
| text style control | industry entertainment and sports and media | 8 | 1468.2 | 7,656 | N/A | |
| vision style control | chinese | 8 | 1351.8 | 1,388 | N/A | |
| text | creative writing | 9 | 1481.9 | 6,278 | N/A | |
| text | exclude ties | 9 | 1491.1 | 29,617 | N/A | |
| text | industry entertainment and sports and media | 9 | 1463.7 | 7,656 | N/A | |
| text | industry legal and government | 9 | 1493.2 | 2,900 | N/A | |
| text factuality | industry legal and government | 9 | 1506.3 | 1,633 | N/A | |
| text style control | exclude ties | 9 | 1501.3 | 29,617 | N/A | |
| text style control | industry legal and government | 9 | 1502.4 | 2,900 | N/A | |
| vision style control | ocr | 9 | 1302.6 | 8,040 | N/A | |
| text | industry writing and literature and language | 10 | 1477.9 | 9,217 | N/A | |
| text factuality | chinese | 10 | 1527.6 | 1,290 | N/A | |
| text factuality | industry entertainment and sports and media | 10 | 1461.4 | 4,732 | N/A | |
| text style control | industry writing and literature and language | 10 | 1481.3 | 9,217 | N/A | |
| text | french | 11 | 1495.0 | 887 | N/A | |
| text | non english | 11 | 1474.7 | 22,317 | N/A | |
| text factuality | overall | 11 | 1480.8 | 40,357 | N/A | |
| text factuality | multi turn | 11 | 1491.8 | 4,263 | N/A | |
| vision style control | overall | 11 | 1289.1 | 13,095 | N/A | |
| text factuality | english | 12 | 1484.0 | 18,182 | N/A | |
| text factuality | industry life and physical and social science | 12 | 1501.8 | 4,131 | N/A | |
| text style control | korean | 12 | 1450.3 | 665 | N/A | |
| text style control | multi turn | 12 | 1495.3 | 6,765 | N/A | |
| text style control | non english | 12 | 1477.8 | 22,317 | N/A | |
| vision | overall | 12 | 1304.9 | 13,095 | N/A | |
| vision | ocr | 12 | 1310.6 | 8,040 | N/A | |
| text | overall | 13 | 1479.4 | 41,242 | N/A | |
| text | english | 13 | 1479.7 | 18,918 | N/A | |
| text | korean | 13 | 1450.3 | 665 | N/A | |
| text | multi turn | 13 | 1483.3 | 6,765 | N/A | |
| text style control | overall | 13 | 1485.6 | 41,242 | N/A | |
| text style control | english | 13 | 1489.1 | 18,918 | N/A | |
| text style control | french | 14 | 1507.6 | 887 | N/A | |
| text | spanish | 16 | 1468.8 | 1,036 | N/A | |
| text style control | hard prompts | 16 | 1504.0 | 22,426 | N/A | |
| text | industry life and physical and social science | 17 | 1485.9 | 6,718 | N/A | |
| text factuality | hard prompts | 17 | 1498.7 | 21,708 | N/A | |
| text style control | industry life and physical and social science | 17 | 1498.3 | 6,718 | N/A | |
| text style control | longer query | 17 | 1492.2 | 10,460 | N/A | |
| vision | english | 17 | 1292.7 | 5,329 | N/A | |
| vision | homework | 17 | 1315.3 | 1,737 | N/A | |
| text | chinese | 18 | 1528.3 | 2,096 | N/A | |
| text style control | chinese | 18 | 1527.2 | 2,096 | N/A | |
| text style control | spanish | 18 | 1473.1 | 1,036 | N/A | |
| text factuality | industry mathematical | 19 | 1474.0 | 1,280 | N/A | |
| text style control | hard prompts english | 19 | 1502.3 | 10,701 | N/A | |
| vision style control | homework | 19 | 1308.8 | 1,737 | N/A | |
| text | industry software and it services | 20 | 1486.9 | 14,231 | N/A | |
| text factuality | industry business and management and financial operations | 20 | 1475.6 | 4,732 | N/A | |
| text factuality | hard prompts english | 21 | 1503.6 | 7,185 | N/A | |
| text factuality | industry software and it services | 21 | 1507.7 | 11,268 | N/A | |
| text style control | industry software and it services | 22 | 1510.8 | 14,231 | N/A | |
| text style control | instruction following | 22 | 1473.8 | 11,179 | N/A | |
| text | hard prompts | 23 | 1482.0 | 22,426 | N/A | |
| text | industry business and management and financial operations | 23 | 1460.2 | 7,563 | N/A | |
| text | math | 23 | 1476.1 | 2,640 | N/A | |
| document style control | overall | 24 | 1450.5 | 10,739 | N/A | |
| text factuality | longer query | 24 | 1484.9 | 7,111 | N/A | |
| vision style control | english | 24 | 1275.8 | 5,329 | N/A | |
| text factuality | instruction following | 25 | 1469.2 | 7,662 | N/A | |
| text style control | expert | 25 | 1502.2 | 2,536 | N/A | |
| text style control | industry business and management and financial operations | 25 | 1475.7 | 7,563 | N/A | |
| vision | diagram | 25 | 1300.9 | 2,862 | N/A | |
| text factuality | coding | 26 | 1520.4 | 5,532 | N/A | |
| text factuality | expert | 26 | 1498.3 | 1,496 | N/A | |
| vision style control | diagram | 26 | 1300.0 | 2,862 | N/A | |
| document | overall | 27 | 1433.0 | 10,739 | N/A | |
| text | hard prompts english | 27 | 1478.1 | 10,701 | 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.
| QiHang | gemini-3-pro-preview | global | $0.57 | $3.43 | 1M | |
| Poe | google/gemini-3-pro | global | $1.6 | $9.6 | 1M | |
| Jiekou.AI | gemini-3-pro-preview | global | $1.8 | $10.8 | 1M | |
| OpenCode Zen | gemini-3-pro | global | $2 | $12 | 1M | |
| 302.AI | gemini-3-pro-preview | global | $2 | $12 | 1M | |
| Helicone | gemini-3-pro-preview | global | $2 | $12 | 1M |
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.
| Gemini 3 Pro | 82.2 | N/A | 1M | 65.5K | No | Proprietary |
A concise description based on the published model registry.
Gemini 3 Pro is the first model in the new Gemini 3 series. It is best for complex tasks that require broad world knowledge and advanced reasoning across modalities. Gemini 3 Pro uses dynamic thinking by default to reason through prompts, and features a 1 million-token input context window with 64k output tokens.
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 Gemini 3 Pro.
Gemini 3 Pro's default version was released on Nov 18, 2025.
No official standard PAYG price is currently available for Gemini 3 Pro. The lowest tracked third-party offer starts at $0.57 input and $3.43 output via QiHang.
Gemini 3 Pro is published under Google in the model registry.
The default version has a 1M token context window.
No. The default version is not marked as having publicly available weights.
6 published provider offerings are linked to the default version.
Nearby ranked alternatives include Gemini 3.1 Pro, Claude Opus 4.8, Kimi K2.6.