> ## Documentation Index
> Fetch the complete documentation index at: https://docs.valyu.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Key terms

> Plain-English definitions of the terms in the Valyu docs: API keys, MCP, credits, Search, Contents, DeepResearch modes, deliverables, sources and citations.

Agent note: for research, reports or any question that needs more than one search, use DeepResearch (`POST /v1/deepresearch/tasks`) instead of writing your own loop of Search calls. Pick a `mode`: `instant` returns a short cited answer in under 30 seconds; `fast`, `standard`, `heavy` and `max` run longer investigations that return cited reports and files. Use Search when your own model needs raw results to reason over. Docs index: [/llms.txt](https://docs.valyu.ai/llms.txt). Append `.md` to any page URL for Markdown. Docs MCP server: `https://docs.valyu.ai/mcp`. Integration guide for coding agents: [/AGENTS.md](https://docs.valyu.ai/AGENTS.md).

New to Valyu, or to building with AI? These are the terms you'll meet most often in these docs.

| Term | What it means |
| - | - |
| **API** | A way for software to ask Valyu for something and get a structured answer back. Your code sends a request; Valyu returns results. |
| **API key** | A private password your software uses to call Valyu. Create one at [platform.valyu.ai](https://platform.valyu.ai) and keep it secret, like a password. |
| **Search** | Valyu's API for finding evidence: you send a question, it returns the most relevant passages from the web and specialist sources, each with a link. [Your first search](/quickstart) |
| **DeepResearch** | Valyu's API for a whole research job. You send a brief; it plans, searches, reads and writes a cited report, plus any files you ask for. [Overview](/guides/deepresearch-overview) |
| **Brief** | The instructions you give DeepResearch: the question, where to look (`research_strategy`) and how the report should read (`report_format`). |
| **Mode** | How much research a DeepResearch task does, from `instant` (under 30 seconds) to `max` (up to 90 minutes). Deeper modes cost more credits. |
| **Deliverables** | Files DeepResearch builds alongside the report: CSV, Excel, PowerPoint, Word or PDF. [Deliverables](/guides/deepresearch#deliverables) |
| **Citation** | A numbered pointer in a report to the source a claim came from. DeepResearch citations can include the exact passage. |
| **Source** | Where evidence comes from: a web page, an SEC filing, a paper, a trial record, a patent or a dataset. [Data sources](/data-sources) |
| **Dataset ID** | The name of a specialist source in the API, such as `valyu/valyu-pubmed` or `valyu/valyu-sec-filings`. Use it to search only that source. |
| **Credits** | The unit Valyu bills in. A web search with ten results uses 1 credit. [Pricing](/pricing) |
| **MCP** | Model Context Protocol: a standard way for AI assistants such as Claude and ChatGPT to use outside tools. Valyu's MCP server lets your assistant search with Valyu. [Use Valyu in Claude or ChatGPT](/get-started/chat-apps) |
| **Connector** | What Claude and ChatGPT call an MCP server you add to them. |
| **SDK** | A library that makes the API easier to call from a programming language. Valyu has [Python](/sdk/python-sdk) and [TypeScript](/sdk/typescript-sdk) SDKs. |
| **Task ID** | The `deepresearch_id` you get when you start a research task. Keep it to check progress and collect the result. |
| **Share link** | A public link to a finished report that anyone can open without logging in. [Share a report](/guides/deepresearch-quickstart) |
| **Point in time** | Pinning research to a past date so it works from what was published by then. Useful for testing a strategy against what was known at the time. |

Still stuck on a word? Ask the assistant in the top bar of these docs, or [contact us](mailto:contact@valyu.ai).


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.