One agent session
1M input tokens and 100K output tokens, no caching.
- Input1M x $2 = $2.00
- Output0.1M x $12 = $1.20
$3.20per session
Google - Closed weights
Gemini 3.1 Pro Preview pricing and specs
Gemini 3.1 Pro Preview is a closed-weights model from Google that costs $2 per 1M input tokens and $12 per 1M output tokens on the first-party API, with a 1.05M-token context window.
Last verified from ai.google.dev. All sources
- Input / 1M
- $2
- Output / 1M
- $12
- Context
- 1.05M tokens
- Weights
- Closed
Pricing per 1M tokens
| Item | Price / 1M tokens |
|---|---|
| InputPrompts up to 200,000 tokens | $2 |
| OutputPrompts up to 200,000 tokens | $12 |
| Cache readPrompts up to 200,000 tokens | $0.20 |
| Batch inputBatch requests, prompts up to 200,000 tokens | $1 |
| Batch outputBatch requests, prompts up to 200,000 tokens | $6 |
| Long-context inputPrompts over 200,000 tokens | $4 |
| Long-context outputPrompts over 200,000 tokens | $18 |
| Long-context cache readPrompts over 200,000 tokens | $0.40 |
| Long-context batch inputBatch requests, prompts over 200,000 tokens | $2 |
| Long-context batch outputBatch requests, prompts over 200,000 tokens | $9 |
Note: Preview model (paid tier only, no free tier). Higher rates apply to prompts over 200k tokens. Context caching storage is $4.50 per 1M tokens per hour rather than a per-write fee, so cache write is not stated; Batch cached input is the same as Standard. Output price includes thinking tokens. Flex equals Batch; Priority is $3.60 / $21.60 / $0.36 cached (<=200k) and $7.20 / $32.40 / $0.72 (>200k). The gemini-3.1-pro-preview-customtools endpoint is priced the same. Release date from the Gemini API changelog (February 19, 2026).
Specs
- Context window
- 1,048,576 tokens
- Max output
- 65,536 tokens
- Release date
- Provider
- API model id
gemini-3.1-pro-preview- Vercel AI Gateway id
google/gemini-3.1-pro-preview- Weights
- Closed weights
What Gemini 3.1 Pro Preview costs in practice
Worked from the standard list rates above, with the math shown so you can swap in your own token counts. These use the base rates, which apply while each request's prompt stays at or under 200,000 tokens; above that, Gemini 3.1 Pro Preview bills $4 input and $18 output for the whole request.
A chat app
10,000 requests a day at 2K input and 500 output tokens each.
- Input20M x $2 = $40.00
- Output5M x $12 = $60.00
$100.00per day
$3,000.00 per 30-day month
The same session at a 90% cache hit
90% of the input is read from the prompt cache. The one-time cache write premium is not included.
- Cached input0.9M x $0.20 = $0.18
- Uncached input0.1M x $2 = $0.20
- Output0.1M x $12 = $1.20
$1.58per session
Compare with
| Model | Input / 1M | Output / 1M |
|---|---|---|
| Gemini 3.1 Pro Preview (this page) | $2 | $12 |
| GPT-5.6 TerraOpenAI | $2 | $12 |
| Amazon Nova PremierAmazon | $2.50 | $12.50 |
| Claude Sonnet 5Anthropic | $2 | $10 |
| Claude Sonnet 5.5Anthropic | $2 | $10 |
| Gemini 2.5 ProGoogle | $1.25 | $10 |
| Gemini 4 ArgonGoogle | $2 | $10 |
| Gemini Robotics ER 2 PreviewGoogle | $1 | $5 |
| Gemini 3.6 FlashGoogle | $0.75 | $3.75 |
| Gemini 3.8 FlashGoogle | $0.75 | $3.75 |
| Gemini 3 Flash PreviewGoogle | $0.50 | $3 |
Compare
Side-by-side pages with both models' prices, specs, cost examples and like-for-like benchmark rows.
Related reading
- Claude Fable 5 vs Gemini 3.1 Pro: July 2026 ComparisonClaude Fable 5 vs Gemini 3.1 Pro: Anthropic's $10/$50 API-only model vs Google's $2/$12 preview on price, context and benchmarks, and how Opus 5 shifts it.
- The Mid-Tier Shootout: GPT-5.4 vs Gemini 3.1 Pro vs DeepSeek V4 ProGPT-5.4 vs Gemini 3.1 Pro vs DeepSeek V4: pricing, benchmarks, context behavior, and license terms for the mid-tier models that carry most production traffic.
- Do Your Agent Hooks Fail Open? Claude Code vs Codex vs Cursor vs Gemini CLIA guard hook that crashes, times out or points at a missing script lets the tool call through by default in Claude Code. v2.1.295 adds onFailure: block. Here is what each coding agent does when your policy hook breaks, with real test runs.
- EmbeddingGemma 2 Guide: Local Multimodal Embeddings for CodeEmbeddingGemma 2 is a 740M open model that embeds code, images, video and audio in one space. Specs, Matryoshka storage math, and where it breaks.
- Google AI Pro vs Gemini API: What Developers Actually GetGoogle AI Pro costs $19.99 a month and has no API key, but includes $10 of monthly Cloud credits. When the subscription wins and when to pay per token.
- How to Keep API Keys Away From Your AI Coding Agent: Claude Code vs Codex vs Cursor vs Gemini CLIA deny rule on .env stops cat but not a Python one-liner. Here are the three layers that actually keep secrets out of a coding agent's context - file rules, environment scrubbing and credential masking - and what Claude Code, Codex, Cursor and Gemini CLI give you for each, with real test runs.
FAQ
How much does Gemini 3.1 Pro Preview cost per million tokens?
Gemini 3.1 Pro Preview costs $2 per 1M input tokens and $12 per 1M output tokens on Google's first-party API, per ai.google.dev on October 17, 2026. Per 1M tokens, cache reads cost $0.20, and the batch rate is $1 input and $6 output. Prompts over 200,000 tokens are billed at $4 input and $18 output for the whole request.
What is Gemini 3.1 Pro Preview's context window?
Gemini 3.1 Pro Preview has a 1,048,576-token context window and returns up to 65,536 output tokens per response.
Are Gemini 3.1 Pro Preview's weights open?
No. Gemini 3.1 Pro Preview is a closed-weights model. You call it through Google's API as gemini-3.1-pro-preview or through Vercel AI Gateway as google/gemini-3.1-pro-preview.
When was Gemini 3.1 Pro Preview released?
Google released Gemini 3.1 Pro Preview on February 19, 2026.
Sources
Last verified from ai.google.dev.
Newsletter
Get model price changes as they ship
One email per week. New model releases, price cuts and what they mean for your bill.