Gemini 4 Carbon Official Specifications, Pricing & Benchmarks
Google DeepMind 4th-gen native omni-modal flagship supporting text, image, video, audio & PDF across 1M+ tokens.
📋 Gemini 4 Carbon Official API & System Card Specifications
Verified against official Google DeepMind developer documentation & pricing
1. Identity & Capabilities
API Spec- API Model ID
- gemini-4-carbon
- Release Date
- 2026-11-10
- 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
- $ 0.1 -95%
- Output / 1M Tokens
- $ 10
- Tiered Pricing Notes
- True native omni-modality (video/audio/PDF/image/text) across 1,048,576 context tokens
🏆 Official Benchmark Results
Sourced from official System Card / Model Card & technical reports
PhD-level Google-proof science reasoning questions
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.
LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code.
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.
A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings.
🎛️ Official Sampling & Reasoning Parameters
Recommended generation settings & API parameter constraints
💡 Keep temperature=1.0 and control reasoning latency via thinking_config.
📊 Gemini 4 Carbon Full Benchmark Scorecard
15 public benchmarks · ranks among 154 models · synced 2026-11-20
💻 Coding 7 items
WebDev Arena is a leaderboard for evaluating AI models on web development tasks, including zero-shot generation, complex prompts, and interactive web UI creation. Models are ranked using Elo ratings based on their performance in coding and web development challenges.
SEC-bench Pro is a self-evolving software-security benchmark that measures agent bug hunting on critical, high-complexity systems.
CWE-Bench evaluates autonomous agents on finding and patching real-world software vulnerabilities.
Artificial Analysis reported accuracy on terminal tasks with executable final-container-state verifiers.
Terminal-Bench-Science 0.1 is a Stanford-led community benchmark of 70 tasks drawn from scientific research workflows across the life, physical, earth, mathematical, and engineering sciences.
FrontierSWE V2 evaluates agents on open-ended technical projects and reports mean@5. Kept separate from FrontierSWE V1 dominance-score results, which are not comparable.
ExploitGym is a large-scale, realistic benchmark built from real-world vulnerabilities across userspace programs, Google's V8 engine, and the Linux kernel.
🤖 Agents & Tools 5 items
Vending-Bench 2 tests longer horizon planning capabilities by evaluating how well AI models can manage a simulated vending machine business over extended periods.
Partial-credit computer-use score on OSWorld 2.1, as reported in the Claude Sonnet 5.5 release. Separate from OSWorld 2.0 and strict completion scores.
Vals Index measures economic impact across finance, coding, legal, and tax work, with every sector weighted by its contribution to U.S. GDP.
Blueprint-Bench 2 is an agentic spatial reasoning benchmark that evaluates a model's ability to understand, plan, and reason over architectural blueprints and other structured spatial documents. Scores are reported as a normalized score.
AutomationBench is a tool-use benchmark that evaluates AI agents on automating real-world workflows, testing their ability to orchestrate tools and complete multi-step automation tasks.
👁️ Multimodal 1 items
Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).
Closest Self-Hostable Open-Weight Alternatives to Gemini 4 Carbon
Open-weight models with full HuggingFace weights for private GPU / Mac deployment
Frequently Asked Questions
Is Gemini 4 Carbon open-source or proprietary?
Gemini 4 Carbon (gemini-4-carbon) 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 4 Carbon?
Gemini 4 Carbon supports 1048K input context and 65,536 max output tokens, priced at $2/1M input ($0.1/1M cached) and $10/1M output.