December 2025 API Roundup: Evals, Vertical Index, New Developer Tooling and More
Welcome to the December You.com API product Roundup—our final update of 2025. This month continues our focus on enabling teams to build reliable, high-trust AI applications on top of You.com.
Across Deep Search, evaluation infrastructure, vertical indexing, new SDKs, and our newly expanded cross-platform integrations, our goal remains simple: make it easier for developers and product teams to bring real-time, web-grounded intelligence into the tools, frameworks, and agentic systems they already use.
As we close the year, we’re also preparing for a major January release that will advance the core capabilities developers rely on—accuracy, latency, freshness, hallucination resistance, and total cost of ownership—marking an important evolution of our search and grounding infrastructure.
Here’s what’s new this month.
Deep Search Early Access Update:Community interest surpassed expectations; rollout underway
Major January Release Teaser: Big enhancements across the full search pipeline
Evals-as-a-Service: Domain-specific evaluation infrastructure for real-world accuracy and reliability insights
Vertical Index: Domain-focused indexing for low-latency, high-precision retrieval
Cross-Platform Integrations: Use You.com APIs inside Vercel AI SDK, OpenAI SDK, and Claude Agent SDK
API Playground: To enable teams go from exploration to real-world prototypes with You.com APIs in minutes
You.com TypeScript SDK: Simpler integration for Node.js and TypeScript developers
Looking Ahead to a Major January Release
We’re preparing a significant January launch focused on improving several pillars that matter deeply to grounding and accuracy in LLM’s and agentic systems: Accuracy, Latency, Freshness, Hallucination resistance, Usability and Total cost of ownership of Agentic workflows.
This upcoming release reflects months of foundational work on our search for LLM’s and agentic workflow infrastructure.
More to come in early January — stay tuned.
Deep Search Early Access Update
As a reminder, Deep Search retrieves live webpages, extracts only the passages that directly answer the query, and verifies that the text genuinely appears on those pages before returning it.
Thank you to everyone who signed up for the Deep Search Early Access program. The level of interest far exceeded expectations, and we ran a small lottery to manage the size of the beta group.
To those in the beta: Thank you! Your feedback will directly shape upcoming iterations.
For everyone else: You’re on the priority list for the next wave of early access programs.
Deep Search continues to evolve quickly, especially around accuracy and deeper multi-step reasoning. We’ll share more early in the new year.
Evals-as-a-Service
We’re introducing a new evaluation service for teams who need to measure and compare search, LLM model and other AI agentic performance in real workflows, not with synthetic benchmarks.
Most existing eval approaches rely on generic datasets and narrow templates that don’t capture the nuance of real-world usage.
What It Does
Our Evals-as-a-Service infrastructure enables domain-specific, workflow-aligned evaluations so you can understand how different models and configurations behave on:
your tasks
your documents
your grounding sources
your latency, budget, and other operational constraints
Why It Matters
If you're working on grounded generation, agent reliability, benchmarking, or model selection, this service gives you a clearer and more realistic picture of your system performance than traditional benchmarks thus enabling reliable, long-term, scalable AI infrastructural and architectural decisions.
For applications that need deeper, more targeted coverage in a specific field, we now support Vertical Indexes—focused, high-precision indexes over curated portions of the web.
What It Does
A Vertical Index allows you to define:
Domains
Competitive landscape
Industry-specific sources
Technical sites, documentation, or research hubs
And more
These indexes span hundreds of thousands of pages and documents and are optimized for low-latency, high-accuracy retrieval.
Why It Matters
This is ideal for applications that need:
Deeper coverage in a specialized field
Predictable relevance
Faster response times
The ability to avoid maintaining custom crawlers or indexing infrastructure
We’ve launched a new Platform Home Page and API Playground to give developers a clearer, faster path into the You.com ecosystem. The new experience brings together your API key, usage, available endpoints, and recommended next steps into a single, clean interface,making it easier to understand what’s possible and how to get started.
What It Does
The updated platform experience includes:
A redesigned platform home with an overview of each API and its purpose
A unified view of your credits, rate limits, and available endpoints
A new API Playground where you can make real API calls using your own key
Native support for Search, Contents, and Agents APIs — no setup required
Inline examples and structured responses for faster iteration
Developers can now explore, test, and prototype with You.com APIs in one place without writing a single line of code.
Why It Matters
Shorter time to first call:The Playground removes friction so developers can try out the you.com APIs and see real outputs immediately
Better platform comprehension: The home page clearly communicates the value and use cases for each endpoint
Faster experimentation: Try queries, tweak parameters, and understand responses without switching tools
Better onboarding for new teammates: Great for PMs, designers, analysts, or engineers new to the platform
A foundation for more interactive tooling: Future updates will introduce richer examples, prebuilt workflows, and guided flows for agents
A Look Ahead
This update is the first step toward a more guided, interactive You.com developer experience.
Over time, the Playground will expand to include:
Multi-step agent demos
Deep search previews
Workflow templates
Vertical Index exploration
Evaluation examples
Our goal is to make the You.com platform feel not just accessible, but immediately useful - helping teams go from exploration to real-world prototypes in minutes. Start building.
Cross-Platform Integrations for Search, Contents, and Agents APIs
We’ve expanded support for You.com’s APIs across several major developer ecosystems, making it easier to use real-time, web-grounded intelligence wherever you build. You can now access our Search, Contents, and Agents APIs through the Vercel AI SDK, the OpenAI SDK, and the Claude Agent SDK.
What It Does
These new integrations let you:
Call You.com APIs directly from Teams bots, Vercel AI workflows, OpenAI agents, and Claude agents
Use the same endpoints and patterns across multiple ecosystems
Add grounded search, content extraction, and agentic reasoning to existing applications without rebuilding your stack
Accelerate prototyping by using tooling and frameworks you already rely on
Minimal setup, no custom plumbing—just plug in your API key and start sending queries.
Why It Matters
Meets developers where they already build: No need to shift environments or rewrite infrastructure
Shortens the path to web-grounded applications: Search, extraction, and agent reasoning become available inside your preferred SDKs
Reduces integration overhead: Consistent behavior across multiple platforms simplifies both prototyping and production rollouts
Enables multi-platform AI strategies: The same You.com grounding layer can now power internal tools, customer-facing apps, and agentic systems across diverse ecosystems
A Look Ahead
This is the beginning of a broader initiative to make You.com’s grounding and agent capabilities portable across the entire AI toolchain. Expect continued expansion into additional SDKs, frameworks, and workflow platforms as we move into 2026.
Wherever developers build AI, You.com’s search infrastructure should be readily available to add accurate, real-time knowledge of the world into their workflows.
We’ve released the first version of our TypeScript SDK, giving JavaScript and Node.js developers a simple, idiomatic way to access You.com’s Search APIs.
What It Does
The TypeScript SDK handles request construction, authentication, response parsing, and error handling — reducing the boilerplate typically required to integrate external APIs.
The SDK provides TypeScript-friendly methods for You.com:
Search
Content and
News APIs
Why It Matters
A consistent SDK experience across Python and TypeScript
Cleaner integrations with backend services, workers, and agent runtimes
Faster prototyping and iteration in JS/TS environments
Strong typing and autocompletion for improved developer ergonomics
As always, we're committed to making advanced AI capabilities more accessible to developers. With new cross-platform integrations, expanded indexing and evaluation infrastructure, and upcoming improvements to our core search capabilities, we’re continuing our mission to deliver the most accurate, extensible, and developer-friendly platform for building search and research for LLMs and agentic applications
Thank you for being part of our API community this year. We’ve appreciated your feedback, ideas, and partnership throughout 2025. We hope you enjoy a restful holiday season, and we’re excited to welcome you into 2026 with a major January release and much more to come.
.webp)
