Gemini 4 Carbon
Gemini 4 Carbon is a chat model from Google, tracked here with independent throughput and latency measurements from Artificial Analysis.
API Pricing
Cheapest input on OpenRouter · cheapest output on OpenRouter| Provider | Input / 1M | Output / 1M | Cached / 1M |
|---|---|---|---|
| $1.00 | $5.00 | $0.050 | |
| $2.00 | $10.00 | $0.100 | |
| $2.00 | $10.00 | $0.100 | |
| $2.00 | $10.00 | - | |
| $2.00 | $10.00 | $0.100 |
Prices updated daily. Last check: Nov 11, 2026
Compare API pricing for every Gemini model →Gemini 4 Carbon pricing by provider
Input, output, and batch rates, plus alternatives, for one provider at a time.
Performance & Benchmarks
Source: Artificial Analysis →Benchmarks measured Nov 2026. Scores are independent evaluations, not vendor-reported.
Model Details
General
- Creator
- Modalities
- Text
Capabilities
Strengths & Limitations
Strengths
- Sustained output throughput measured at about 106 tokens per second by Artificial Analysis, a normal streaming rate for interactive reading
- Part of Google's Gemini line, so it is generally reachable through the same API surfaces and SDKs used for other Gemini models
- Latency profile (long pre-generation delay, fast streaming) is consistent with extended internal deliberation before answering
- Available from multiple serving routes, letting buyers compare rates in the pricing table on this page
- Independent third-party performance measurement exists for the model rather than vendor-reported figures only
- Slow-start-fast-finish behavior is predictable enough to plan batch job scheduling around
Limitations
- Time to first token measured at roughly 124 seconds, which rules out real-time chat, voice, and autocomplete use
- We do not currently track a confirmed context window size for this model
- No accuracy or reasoning benchmark scores are recorded in our database for Gemini 4 Carbon
- Supported input and output modalities are not confirmed in our data, so multimodal plans should be verified against Google's documentation
- Long per-request wall-clock time makes high-volume synchronous workloads expensive in engineering complexity even where token rates are reasonable
Key Features
About Gemini 4 Carbon
Common Use Cases
Gemini 4 Carbon suits asynchronous and batch workloads where a long wait before the first token is tolerable: offline document analysis, overnight code review passes, research summarization queues, data labeling and enrichment jobs, and agent steps that run in the background rather than in front of a waiting user. Because generation streams at roughly 106 tokens per second once it starts, long-form outputs still complete in reasonable time after the initial delay, which makes it workable for report drafting and extended written analysis submitted through a job queue. It is a poor fit for customer-facing chat, voice assistants, IDE autocomplete, or any interface where a user is watching a cursor — for those, pick a model with a sub-second time to first token. Teams building agent pipelines should also budget the 124-second startup cost per call, since multi-step chains multiply it.
Frequently Asked Questions
How much does Gemini 4 Carbon cost?
Pricing varies by provider and by pricing type — input tokens, output tokens, and any cached or batch rates are often billed differently, and providers change rates frequently. Check the pricing table on this page for the current figures across the providers we track.
What is Gemini 4 Carbon best used for?
Asynchronous and batch text work: document analysis, long-form drafting, background agent steps, and research summarization. Its measured ~124-second time to first token makes it unsuitable for live chat or any latency-sensitive interface.
Why does Gemini 4 Carbon take so long to respond?
Artificial Analysis measures a time to first token of roughly 124 seconds, meaning the model spends a long period processing before any output appears. That pattern is typical of models that run extended internal deliberation. Once generation starts, output streams at about 106 tokens per second.
What context window does Gemini 4 Carbon support?
We do not have a confirmed context window figure for Gemini 4 Carbon in our database. Check Google's model documentation or your chosen provider's API reference for the supported input length before designing around a specific limit.
Can Gemini 4 Carbon accept images or other non-text input?
Our data does not confirm the input modalities for this model, so we do not list it as text-only or multimodal. Verify supported input types with Google's model card or the provider you plan to use.
Is Gemini 4 Carbon a good choice for a customer-facing chatbot?
Generally no. With a measured time to first token near two minutes, users would see nothing on screen for an extended period. A model with low first-token latency is a better match for interactive chat; reserve Gemini 4 Carbon for work that runs out of the user's view.