Google launches Gemini Decisions API powered by Gemini Decision 1

Executive Briefing
“Google has introduced the Gemini Decisions API powered by Gemini Decision 1, targeting...”
Key Takeaways
- 1.Gemini Decision 1 does not generate text, instead scoring candidate options in a single forward pass to return calibrated probabilities.
- 2.Google prices the service at $0.10 per one million input tokens with zero output tokens billed as usage remains zero.
- 3.The model supports a 262,144-token context and natively reads text, images, video, audio, and PDFs in one prefill pass.
Google has introduced the Gemini Decisions API powered by Gemini Decision 1, targeting branching calls inside automated workflows such as routing, policy checks, and tool selection. The interface scores up to 1,000 typed questions against a shared input at a price of $0.10 per one million input tokens.
Key Developments and Technical Context
Gemini Decision 1 does not generate text, instead scoring candidate options in a single forward pass to return calibrated probabilities.
Google prices the service at $0.10 per one million input tokens with zero output tokens billed as usage remains zero.
The model supports a 262,144-token context and natively reads text, images, video, audio, and PDFs in one prefill pass.
Strategic Analysis and Industry Implications
Eliminating output tokens and billing shared input once across numerous questions helps developers optimize costs when executing repeated branching calls within agentic systems.
Outlook and Key Factors to Watch
The actual performance of the model in large-scale production environments beyond Google's self-reported benchmark tables remains to be independently verified.







