Introducing Tool Retrieval in the Gemini API
Tool Retrieval is now in preview in the Gemini Interactions API. Learn practical implementation patterns and deployment details at ZeroLabs.
What was announced

Google AI Blog released an official announcement regarding Introducing Tool Retrieval in the Gemini API on November 19, 2026.
Tool Retrieval is now in preview in the Gemini Interactions API. Load tool parameter schemas on demand instead of sending your full tool library in every prompt,designed to reduce input tokens and keep tool selection accurate as your agent scales. Product Manager, Gemini API Research Scientist, Google DeepMind We are launching Tool Retrieval in...
Key technical developments and highlights detailed in the release include:
- Explore other styles:: Today we're launching Tool Retrieval in the Gemini API, available in preview in the Interactions API . When an agent connects to dozens or even hundreds of tools, sending every full parameter schema on every request adds up fast. It inflates input token counts, adds latency,...
- How it works: To turn on Tool Retrieval, add a tool_search tool to your request and set defer_loading: true on the tools you want to hold back. The model still sees the name and description of every deferred tool, so it knows what's available, but the parameter schemas stay out of context....
- How client-side retrieval works: If you maintain your own tool registry, embedding index, or per-user permission filter, declare the tool_search with "execution": "client" and define your search parameters. Client-side retrieval adds one round trip per search, because your app runs the lookup. Tools do not...
- How pricing and supported models work: Tool Retrieval has no separate charge. Input and output tokens are billed at the standard rates for the model you call. Turning on tool_search adds a small, fixed block of input tokens, and each schema the model loads is billed as input for the rest of the conversation, the...
- How to get started: Read the documentation - Full setup guide with examples Compare Gemini models - Find the best model for your use case Join the discussion on Discord - Share feedback with the team Check your inbox to confirm your subscription.
Why this matters for technical teams
Updates across foundation models and developer ecosystems directly affect platform baselines, integration ergonomics, and operational trade-offs.
- Developer ergonomics & scaffolding: New features alter baseline expectations and test whether custom agent glue code can be simplified.
- Operational trade-offs: Upgrades to foundation models redefine reasoning ceilings, token latency, and autonomous tool calling reliability across agent architectures.
- Evaluation priorities: Builders should benchmark this release against their domain-specific evaluation harnesses to measure real-world task accuracy and latency.
Source documentation and primary references: Google AI Blog.
ZeroLabs Take
Real architectural progress matters far more than synthetic benchmark vanity scores. While Google AI Blog's release notes for Introducing Tool Retrieval in the Gemini API showcase impressive capability figures, builders care about three concrete realities: function-calling determinism, end-to-end token latency, and predictable operational cost. If the updates documented in this release deliver verifiable improvements in multi-step task execution without degrading instruction following, it represents genuine engineering value. We welcome capability jumps that reduce brittle agent glue code, but recommend evaluating them on your own task distributions before refactoring production systems.
What ZeroLabs is watching next
- Community stress tests: Tracking independent evaluations of instruction following, needle-in-a-haystack retrieval, and complex coding harnesses.
- Tool-calling determinism: Measuring structured output schema compliance under load across complex agent swarms.
- Frontier competition: Watching how Anthropic, Google, and OpenAI adjust their matching latency and pricing tiers in response.
FAQ
- What was officially announced?
Google AI Blog published details on Introducing Tool Retrieval in the Gemini API, focusing on frontier model architecture and developer capabilities.
- How does this affect existing developers and users?
Builders should benchmark this release against their domain-specific evaluation harnesses to measure real-world task accuracy and latency.
- Where can I access the complete release details?
The full announcement and documentation are available directly from Google AI Blog.