Google · released Feb 19, 2026
Gemini 3.1 Pro Preview
Gemini 3.1 Pro Preview is Google's older flagship, and Gemini 4 Argon beats it for less money.
8 months oldThinks before answering
- Input price
- $2 / M
- Output price
- $12 / M
- Context
- 1M tokens
- Longest reply
- 66K
- Reads
- text, image, video, audio, pdf
- Knows up to
- 2025-01
Where it shines
- It scores 73.5 on SimpleQA.
- It scores 1482 on creative writing.
Where it struggles
- Its document arena score of 1444 is weak.
Where you can use it
- Gemini APIAPI$2 in / $12 out per million tokens.
- OpenRouterAPIListed as
google/gemini-3.1-pro-preview
How it scores on each job
Scores run 0 to 100 and blend every measure we have for that job. The number after the dot is its rank among the 33 models in the lineup.
The evidence
| Measure | Score | Rank in lineup | Source |
|---|---|---|---|
| Epoch Capabilities Index | 155 | #21 | Epoch AI |
| LMArena text | 1480 | #10 | LMArena |
| LMArena hard prompts | 1485 | #16 | LMArena |
| LMArena instruction following | 1466 | #18 | LMArena |
| LMArena multi-turn | 1487 | #10 | LMArena |
| LMArena creative writing | 1482 | #7 | LMArena |
| LMArena writing & language | 1481 | #7 | LMArena |
| LMArena coding | 1484 | #19 | LMArena |
| LMArena math | 1486 | #16 | LMArena |
| LMArena business & finance | 1459 | #17 | LMArena |
| LMArena legal & government | 1486 | #12 | LMArena |
| LMArena medicine | 1487 | #11 | LMArena |
| WebDev Arena | 1447 | #26 | LMArena |
| Vision Arena | 1296 | #9 | LMArena |
| Document Arena | 1444 | #8 | LMArena |
| Agent Arena | #35 | #20 | LMArena |
| SWE-bench Verified | 76% | #4 | Epoch AI |
| Terminal-Bench | 80% | #3 | Epoch AI |
| GPQA Diamond | 94% | #7 | Epoch AI |
| Humanity's Last Exam | 46% | #3 | Epoch AI |
| FrontierMath | 60% | #23 | Epoch AI |
| SimpleQA Verified | 74% | #4 | Epoch AI |
| METR time horizon | 6.4 h | #1 | Epoch AI |
| ARC-AGI-2 | 77% | #11 | Epoch AI |
| DeepResearch Bench | 48 | #2 | Epoch AI |
Compare it with these
8 weeks old
Gemini 3.8 Flash
Gemini 3.8 Flash gives near-flagship answers for a small fraction of the price.
Everyday questionsResearchImages & screenshots
- Price
- $0.75 in · $3.75 out
- Context
- 1M
3 months old
Gemini 3.7 Flash
Gemini 3.7 Flash is a cheap, capable Google model that Gemini 3.8 Flash has replaced.
WritingEveryday questions
- Price
- $0.75 in · $3.75 out
- Context
- 1M
NewOpen weights
Meta
Muse Spark Open
Muse Spark Open is Meta's new open-weight coding model, and early results put it mid-pack.
Research
- Price
- $1.25 in · $4.25 out
- Context
- 1M