Google · Released Feb 19, 2026Gemini 3.1 Pro Preview
Reasoning-first Gemini preview for agentic coding and complex problem solving.
- Capability
- 154.8ECI · #31 of 148
- Input
- $2.00per 1M tokens
- Output
- $12.00per 1M tokens
- Context
- 1.05M66K max output
What it costs
- Input
- $2.00per million tokens
- Output
- $12.00per million tokens
- Cached input
- $0.20per million tokens
- Over 200K tokens
- $4.00 / $18.00input / output per million
- Blended (3:1)
- $4.50pricier than 86% of priced models
Typical monthly bills
| Monthly usage | Estimated cost |
|---|---|
| Side project2M input + 0.5M output tokens | $10.00 |
| Team assistant25M input + 5M output tokens | $110.00 |
| Production app250M input + 50M output tokens | $1,100 |
Official Google API price, as listed on models.dev.
Independent benchmarks
#31 of 148 scored models
The shaded band is Epoch AI’s confidence range (152.6–157.3); the tick marks the median scored model.
Scores from Epoch AI, run independently of Google.
The details
- Lab
- Released
- Feb 19, 2026
- Knowledge cutoff
- Jan 2025
- Context window
- 1,048,576 tokens
- Max output
- 65,536 tokens
- Inputs
- Text, Images, PDFs, Audio, Video
- Output
- Text
- Reasoning
- Adjustable effort low · medium · high
- Tool calling
- Yes
- Structured output
- Yes
- Weights
- Proprietary
- API model ID
gemini-3.1-pro-previewon Google- Availability
- 26 API providerslisted on models.dev
What Google claims
Published by the lab at launch. Settings vary, so compare these only with care.
| Benchmark | Score | Setting | Source |
|---|---|---|---|
| SWE-Bench Pro | 54.2 resolve rate | — | Source |
| Terminal-Bench v2.1 | 70.3 success rate | — | Source |
| SWE-Bench Pro | 46.1 resolve rate | — | Source |
| SWE-Atlas Codebase QnA | 13.5 | — | Source |
| SWE-Atlas Refactoring | 33.81 | — | Source |
| SWE-Atlas Test Writing | 29.84 | — | Source |
| Artificial Analysis Coding Agent Index | 43 | high | Source |
| SWE-Atlas Codebase QnA | 45.6 | high | Source |
| SWE-Bench Pro | 15.1 | high | Source |
| Terminal-Bench v2.1 | 68.3 | high | Source |
| GPQA Diamond | 94.3 | — | Source |
| Humanity's Last Exam | 44.4 | — | Source |
More from Google
About Gemini 3.1 Pro Preview
How much does Gemini 3.1 Pro Preview cost?
Gemini 3.1 Pro Preview costs $2.00 input / $12.00 output per million tokens (official Google API price). Cached input is $0.20 per million tokens. Requests over 200K tokens are billed at $4.00 input / $18.00 output. At a 3:1 input-to-output mix that is $4.50 per million tokens, more expensive than 86% of the 360 priced models we track.
What is the context window of Gemini 3.1 Pro Preview?
Gemini 3.1 Pro Preview accepts up to 1,048,576 tokens per request and can write up to 65,536 tokens in one response.
How good is Gemini 3.1 Pro Preview?
Epoch AI gives Gemini 3.1 Pro Preview a Capabilities Index score of 154.8 (likely range 152.6–157.3), ranking it #31 of 148 models Epoch has scored. Epoch AI benchmark results: GPQA Diamond 94.4%, FrontierMath Tiers 1–3 59.7%, OTIS Mock AIME 2024–2025 95.6%, SWE-bench Verified 75.6%, SimpleQA Verified 73.5%.
Is Gemini 3.1 Pro Preview open source?
No. Gemini 3.1 Pro Preview is proprietary; you use it through Google’s API or partner platforms.
What inputs does Gemini 3.1 Pro Preview support?
Gemini 3.1 Pro Preview accepts text, images, PDFs, audio and video and replies in text. It is a reasoning model with low, medium and high effort settings, supports tool calling and can return structured JSON output.
When was Gemini 3.1 Pro Preview released?
Google released Gemini 3.1 Pro Preview on Feb 19, 2026. Its training data runs to Jan 2025.
What are the best alternatives to Gemini 3.1 Pro Preview?
The closest current models from other labs on capability, price and release date are GPT-5.3 Codex (OpenAI, ECI 156.8, $1.75 / $14.00), Qwen3.7 Max (Alibaba (Qwen), ECI 153.7, $2.50 / $7.50), Claude Sonnet 4.6 (Anthropic, ECI 152.3, $3.00 / $15.00) and Grok 4.5 (xAI, ECI 154.0, $2.00 / $6.00).