ai.canshu.com
中文
VRAM Calculator Advisor & TCO API Pricing Compare Models Leaderboard GPUs

Search models, GPUs, and tools

↑↓ Navigate · Enter Open · ★ Watchlist prioritized ai.canshu.com
Google DeepMind
Proprietary API Google DeepMind · gemini-3.1-pro · Text+Image+Video+Audio+PDF

Gemini 3.1 Pro Official Specifications, Pricing & Benchmarks

Early-2026 Google multimodal flagship achieving 94.3% on GPQA Diamond and 91.0% on MMLU-Pro.

📋 Gemini 3.1 Pro Official API & System Card Specifications

Verified against official Google DeepMind developer documentation & pricing

API Model ID: gemini-3.1-pro

1. Identity & Capabilities

API Spec
API Model ID
gemini-3.1-pro
Release Date
2026-02-19
Weight License
Proprietary (Undisclosed)
Native Reasoning
✅ Supported
Function Calling & JSON Schema
✅ 原生支持
Batch API 异步批处理
✅ Supported (50% off)

2. Context, Output & Modalities

Limits
Max Context Window
1,048,000 tokens (1048K)
Max Output Tokens
65,536 tokens
Input Modalities
Text + Image + Video + Audio + PDF
Output Modalities
Text
Direct China Access
🌐 Global / OpenRouter
Deployment Mode
Cloud API only (Proprietary)

3. Official Pricing & Cache

USD / 1M
Input / 1M Tokens
$ 2
Cached Input / 1M
$ 0.2 -90%
Output / 1M Tokens
$ 12
Tiered Pricing Notes
≤200K ($2/$12); >200K ($4/$18); scored 94.3% on GPQA Diamond

🏆 Official Benchmark Results

Sourced from official System Card / Model Card & technical reports

Arena Elo ~1514 AA Index 54
🧠 Reasoning
GPQA Diamond 94.3 %

PhD-level Google-proof science reasoning questions

🧠 Reasoning
AIME 2026 98.3 %

All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999.

📚 Knowledge
MMLU-Pro 91 %

A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks.

💻 Coding
LiveCodeBench v6 91.7 %

LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code.

💻 Coding
SWE-bench Verified 80.6 %

A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.

🤖 Agentic
BrowseComp 85.9 %

BrowseComp is a benchmark comprising 1,266 questions that challenge AI agents to persistently navigate the internet in search of hard-to-find, entangled information.

🎛️ Official Sampling & Reasoning Parameters

Recommended generation settings & API parameter constraints

Thinking High Mode
temperature=1 top_p=0.95 max_tokens=65,536

📊 Gemini 3.1 Pro Full Benchmark Scorecard

22 public benchmarks · ranks among 154 models · synced 2026-11-20

⚔️ Compare head-to-head

💻 Coding 8 items

LiveCodeBench Pro #1/4 2887

LiveCodeBench Pro is an advanced evaluation benchmark for large language models for code that uses Elo ratings to rank models based on their performance on coding tasks.

SWE-Bench Verified #7/59 80.6%

A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.

MCP Atlas #25/39 69.2%

MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.

Terminal-Bench 2.0 #11/38 68.5%

Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal.

SciCode #3/24 59%

SciCode is a research coding benchmark curated by scientists that challenges language models to code solutions for scientific problems.

SWE-Bench Pro #41/52 54.2%

SWE-Bench Pro is an advanced version of SWE-Bench that evaluates language models on complex, real-world software engineering tasks requiring extended reasoning and multi-step problem solving.

FrontierSWE #10/16 40%

FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. Performance is reported as a dominance score, where higher is better.

DeepSWE 1.1 #48/49 12%

DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.

Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).

Closest Self-Hostable Open-Weight Alternatives to Gemini 3.1 Pro

Open-weight models with full HuggingFace weights for private GPU / Mac deployment

Compare with these →

Frequently Asked Questions

Is Gemini 3.1 Pro open-source or proprietary?

Gemini 3.1 Pro (gemini-3.1-pro) is a proprietary commercial model by Google DeepMind. Weights and internal layer dimensions are undisclosed; it is accessed via cloud API.

What are the context limit and API pricing of Gemini 3.1 Pro?

Gemini 3.1 Pro supports 1048K input context and 65,536 max output tokens, priced at $2/1M input ($0.2/1M cached) and $12/1M output.