How I filter agent tools without burning output tokens
Running routing, policy checks, or ranking through a generative model means paying for output text you never wanted. Gemini Decision 1 on Vercel's AI Gateway returns booleans, choices, and scores instead, avoiding that output entirely.
What it is
Gemini Decision 1 is a multimodal decision model from Google, accessible through AI Gateway with the AI SDK, TypeSafe-compatible API, or Decisions API. It takes text, images, audio, video, and PDFs as input and returns structured decisions rather than generated text.
Applied case
A developer building an agent can use the model to filter which tools an incoming request is allowed to call, asking a Boolean question about each tool before execution. The same pattern works for policy checks on user-submitted PDFs or ranking retrieved results by score. Decision calls show up in AI Gateway logs and count toward budgets like any other model request, so usage and cost stay visible. PDF, audio, and video inputs require AI SDK 7, while the TypeSafe-compatible and Decisions APIs support text and image inputs.
- Use google/gemini-decision-1 via the AI SDK, TypeSafe API, or Decisions API for tool filtering and ranking
When to use it
Use when you need routing, policy checks, tool filtering, or ranking as booleans, choices, or scores.
Skip when you need generated text or structured content output.
Read more
Recommended for you
- WorkflowVercel
When to use AI Gateway instead of direct image model APIs?
A separate MiniMax account, key, and bill just to generate product images slows you down until one gateway API key covers it.
- WorkflowVercel
When to use Muse Ember instead of a big reasoning model?
Agents repeat bounded sub-tasks like tool calls many times per run, and paying heavyweight-model prices for each one drains budget and latency.
- WorkflowVercel
When to use AI Gateway instead of calling model providers directly?
Scattered keys, per-provider billing, and outages on any single provider make direct model calls fragile; AI Gateway fixes this at no markup.
- WorkflowVercel
When to use Token Pass instead of per-model API bills?
Separate provider accounts and surprise API bills for each coding agent quietly drain team budgets; one subscription now covers the lineup.
Enjoyed this article?
Subscribe for new articles. No spam. Unsubscribe anytime.