Gemini 4 Carbon is a frontier Gemini 4 model from Google for coding, agentic and complex reasoning work. It takes text, image, audio, video and file input, supports long decode of up to 1M output tokens through continuation, and controls thinking depth with low, medium and high reasoning effort.
Providers
Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick: Balanced (price + speed), Nitro (fastest), or Floor (cheapest).
50% off | $2.00$1.00 | $10.00$5.00 | $0.10$0.05 | $0.10$0.05 | -- | -- |
|---|
Input $1.00 / Output $5.0050% off-- tok/s-- uptime
Pricing
The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.
Weighted average
Gemini 4 Carbon (batch) — Price history
| $0.00 | $0.00 | 4.8% | 100.0% |
- Effective in $0.00 / Effective out $0.00Token share100.0%
Performance
Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).
Uptime
Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.
Not enough uptime data to display yet.
When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and provider routing options.
Benchmarks
Scores on standardized evaluations and this model’s own outputs on shared benchmark prompts. Higher percentages are better — and each rank shows where this model lands among the models evaluated on that source.
| Source | Benchmark | Score |
|---|---|---|
| Artificial Analysis | Gemini 4 Carbon (High) Intelligence Index | 60.0 |
| Artificial Analysis | Gemini 4 Carbon (High) AA-LCR | 79.3% |
| Artificial Analysis | Gemini 4 Carbon (High) GDPval-AA | 61.6% |
| Artificial Analysis | Gemini 4 Carbon (High) Terminal-Bench 4.0 | 65.2% |
| Artificial Analysis | Gemini 4 Carbon (High) AA-Omniscience Accuracy | 58.3% |
| Artificial Analysis | Gemini 4 Carbon (High) AA-Omniscience Non-Hallucination Rate | 88.6% |
| Design Arena | Gemini 4 Carbon Agents Arena Mobile Apps Elo | 1301 |
| Design Arena | Gemini 4 Carbon Agents Arena Pdf Elo | 1310 |
| Design Arena | Gemini 4 Carbon Models Arena 3D Elo | 1429 |
| Design Arena | Gemini 4 Carbon Models Arena Code Categories Elo | 1354 |
| Design Arena | Gemini 4 Carbon Models Arena Game Development Elo | 1319 |
| Design Arena | Gemini 4 Carbon Models Arena UI Component Elo | 1346 |
| Design Arena | Gemini 4 Carbon Models Arena Website Elo | 1334 |
| Vals AI | Gemini 4 Carbon EMB v1.1 | 76.5% |
| Vals AI | Gemini 4 Carbon Finance Agent v2.1 | 67.2% |
| Vals AI | Gemini 4 Carbon Vals Finance Index | 67.0% |
| Vals AI | Gemini 4 Carbon HLAB-Vals-1.1 | 17.5% |
| Vals AI | Gemini 4 Carbon Vals Legal Index | 75.9% |
| Vals AI | Gemini 4 Carbon Legal Research Bench | 55.8% |
| Vals AI | Gemini 4 Carbon Vals RSI Index | 27.7% |
| Vals AI | Gemini 4 Carbon Tax Agent Bench | 51.8% |
| Vals AI | Gemini 4 Carbon Terminal-Bench 4.0 | 63.1% |
| Vals AI | Gemini 4 Carbon Vals Index | 70.5% |
| Vals AI | Gemini 4 Carbon Vibe Code Bench v1.1 | 89.1% |
Apps
Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.
Not enough data for apps using this model yet.
Activity
Token volume and request traffic to this model over time.
API
Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.
Not currently available through EU in-region routing. In-region routing docs
Frequently asked questions
Gemini 4 Carbon is a frontier Gemini 4 model from Google for coding, agentic and complex reasoning work. It takes text, image, audio, video and file input, supports long decode of up to 1M output tokens through continuation, and controls thinking depth with low, medium and high reasoning effort.
Gemini 4 Carbon (batch) costs $1.00/M input tokens and $5.00/M output tokens, with separate rates for Cache Read at $0.05/M tokens, Cache Write at $0.1875/M tokens, Image Input at $1.00/M tokens, Input Audio at $1.00/M tokens, Input Audio Cache at $0.05/M tokens, and Web Search at $14.00/1K calls.
Gemini 4 Carbon (batch) has a 1,048,576 token context window. It supports up to 1,048,576 completion tokens.
Yes. Gemini 4 Carbon (batch) accepts tools and tool_choice for function calling. It also supports structured outputs via a JSON schema in response_format.
Gemini 4 Carbon (batch) accepts audio, files such as PDFs, images, text, and video as input and returns text.
Gemini 4 Carbon (batch) is served by 1 provider on OpenRouter: Google Vertex.
Gemini 4 Hydrogen, Gemini Decision 1, Gemini Omni 1.2 Flash and 31 more are other text models from Google.
Gemini 4 Carbon (batch) was released on November 10, 2026.