Drafted with AI tools and edited by a human. The figures and conclusions were checked by a GPUniq editor.
TL;DR
The Gemini 4 Hydrogen API costs $1.50 per million input tokens and $7.50 per million output tokens. It features a 1,048,576-token context window and adjustable reasoning levels scoring up to 54 on the Artificial Analysis Intelligence Index. Google AI Studio provides direct access.
What is the API price?
Google DeepMind prices input tokens at $1.50 per million. Output tokens including thinking cost $7.50 per million. These rates stay flat across the entire context window.
How large is the context window?
The unified context window holds 1,048,576 tokens. You can ingest massive documents, codebases, or video files in a single prompt.
The foreground output cap sits at 65,536 tokens per segment. Longer tasks require continuation calls.
| Metric | Specification |
|---|---|
| Unified Context Window, tokens | 1048576 |
| Foreground Output Cap, tokens | 65536 |
| Input Price, $/million tokens | 1.50 |
| Output Price, $/million tokens | 7.50 |
What reasoning levels are available?
Four reasoning controls exist: low, medium, high, and xhigh. High serves as the default setting.
The Artificial Analysis Intelligence Index score hits 51 at default high. Bumping the configuration to xhigh pushes the score to 54.
How does it compare to Gemini 3.8 Flash?
Gemini 4 Hydrogen succeeds Gemini 3.8 Flash. Image-to-Text Safety improves by 7.00 percentage points.
| Feature | Gemini 4 Hydrogen | Gemini 3.8 Flash |
|---|---|---|
| Image-to-Text Safety, gain | +7.00 percentage points | Baseline |
| Context Window, tokens | 1048576 | Not published |
| Max Output Tokens | 65536 | Not published |
| Reasoning Controls | Four adjustable levels | Not published |
Where can developers access the API?
Google AI Studio opened access to the Gemini API on November 12, 2026. You can also integrate via the Gen AI SDK. Enterprise accounts manage rate limits through Google Cloud.
Related on the GPUniq blog
Methodology
Facts are compiled from the vendor's own announcement and documentation, linked below, at the time of writing; where the vendor hasn't published a number it is marked as such. GPUniq prices, when the model is already available here, come from our live catalog.
Sources
- 1.Gemini 4 Hydrogen — Model Card — Google DeepMind (accessed )
- 2.Introducing Gemini 4 Hydrogen for Agentic Workflows — Google (accessed )
Want cheap GPUs for your next project?
Browse live GPU prices and rent the right card in seconds — H100, A100, RTX 4090, and 50+ more models.
