docs.perplexity.ai llms.txt
In short
docs.perplexity.ai serves a 42 KB llms.txt with 2 H2 sections and 190 links, graded A on the llmstxt.org structure (6 of 6 checks). It passes every check. The home page links the file. Read 4 November 2026. The file is one of 2,529 plain-text llms.txt files in this edition of the index.
llms.txt 42 KB, 201 lines
# Perplexity
> Perplexity API docs. Search: ranked web results with extracted page text for your own model. Agent: hosted models with web search and tools, returning a cited answer. Decisions: calibrated probabilities. Orchestrator (private preview): OpenAI- and Anthropic-format routing. Embeddings: vectors.
- [Overview](https://docs.perplexity.ai/docs/getting-started/overview.md): Perplexity API Platform - build with the Orchestrator, Agent, Search, Embeddings, and Decisions APIs. Real-time, web-wide research and Q&A capabilities for your products.
- [Quickstart](https://docs.perplexity.ai/docs/getting-started/quickstart.md): Generate an API key and make your first call in < 3 minutes.
- [Pricing](https://docs.perplexity.ai/docs/getting-started/pricing.md): Understand Perplexity API pricing, model rates, request fees, billing setup, and rate-limit references.
- [Perplexity API MCP Server](https://docs.perplexity.ai/docs/getting-started/integrations/mcp-server.md): Connect AI assistants to Perplexity's search and reasoning capabilities using the Model Context Protocol (MCP).
- [Perplexity Computer MCP Server](https://docs.perplexity.ai/docs/getting-started/integrations/computer-mcp-server.md): Give AI assistants the ability to complete real-world tasks using Perplexity Computer through the Model Context Protocol (MCP).
- [Perplexity CLI](https://docs.perplexity.ai/docs/cli/overview.md): Install and use Perplexity's pplx command-line interface for web search and query-relevant page snippets from your terminal or coding agent.
- [Perplexity Orchestrator API](https://docs.perplexity.ai/docs/router/quickstart.md): Unified access to open-weight models through OpenAI Chat Completions, OpenAI Responses, and Anthropic Messages APIs, with your Perplexity API key.
- [Orchestrator Models & Pricing](https://docs.perplexity.ai/docs/router/models.md): The model catalog for the Perplexity Orchestrator API, with per-token pricing for Perplexity-hosted and third-party models.
- [Routing & Reliability](https://docs.perplexity.ai/docs/router/routing-and-reliability.md): How the Perplexity Orchestrator API routes requests across model deployments, and what happens automatically when a provider degrades or fails.
- [Agent API](https://docs.perplexity.ai/docs/agent-api/quickstart.md): The Agent API is a multi-provider, interoperable API specification for building LLM applications. Access models from multiple providers with integrated real-time web search, tool configuration, reasoning control, and token budgets[…]all through one unified interface.
- [Define the run](https://docs.perplexity.ai/docs/agent-api/building-agents/define-the-run.md): Configure the agent run: start from an effort level, then customize the system prompt, the loop bound, and the model.
- [Prompt the agent](https://docs.perplexity.ai/docs/agent-api/building-agents/prompt-the-agent.md): Split the standing rules (the system prompt) from the question (input), and ground the run so the agent searches and answers on the right thing.
- [Give it tools](https://docs.perplexity.ai/docs/agent-api/building-agents/give-it-tools.md): Enable built-in tools and your own custom tools in the tools array, let the model decide when to call them, and read the results back from the response.
- [Run code](https://docs.perplexity.ai/docs/agent-api/building-agents/run-code.md): Give the agent a sandbox so it computes exact answers instead of estimating them, runs real code, and returns generated files.
- [Structure the output](https://docs.perplexity.ai/docs/agent-api/building-agents/shape-output.md): Force the model to return JSON your code can parse instead of prose you have to scrape.
- [Keep context](https://docs.perplexity.ai/docs/agent-api/building-agents/keep-context.md): Carry a conversation across turns by replaying it in the input array, so a follow-up understands what came before.
- [Agent API Models](https://docs.perplexity.ai/docs/agent-api/models.md): Compare Agent API models, token pricing, supported service tiers, and provider documentation.
- [Effort](https://docs.perplexity.ai/docs/agent-api/effort.md): Explore Perplexity's Agent API effort levels - pre-configured setups optimized for different use cases with specific models, search configs, and tool access.
- [Profiles](https://docs.perplexity.ai/docs/agent-api/profiles.md): Run an Agent API request with a reusable, versioned configuration that you save and manage.
- [Model Fallback](https://docs.perplexity.ai/docs/agent-api/model-fallback.md): Specify multiple models in a fallback chain for higher availability and automatic failover.
- [Tools overview](https://docs.perplexity.ai/docs/agent-api/tools/overview.md): The tools an Agent API run can call - Perplexity-hosted built-in tools, remote MCP servers, and your own custom functions - and where to go for each.
- [Web Search](https://docs.perplexity.ai/docs/agent-api/tools/web-search.md): Search the web from the Agent API with filters, search configurations, pricing, parameters, and response fields.
- [Image Search](https://docs.perplexity.ai/docs/agent-api/tools/image-search.md): Search for web images from the Agent API with result limits, domain and format filters, typed results, streaming events, and per-invocation pricing.
- [Sandbox](https://docs.perplexity.ai/docs/agent-api/tools/sandbox.md): Execute code in isolated containers from inside an Agent API request.
First 28 of 201 lines. 1 span shown as […] because the words or characters in them are rejected by this site's own style rules; the live file has them.
Parse graded A, 6 of 6
file: https://docs.perplexity.ai/llms.txt
Perplexity
Perplexity API docs. Search: ranked web results with extracted page text for your own model. Agent: hosted models with web search and tools, returning a cited answer. Decisions: calibrated probabilities. Orchestrator (private preview): OpenAI- and Anthropic-format routing. Embeddings: vectors.
2 sections, 190 links, an Optional section.
Ora's cached score: 76, grade B, scanned 24 August 2026.
The parse
docs.perplexity.ai lists 190 links: 190 point at perplexity.ai or a relative path, 0 point elsewhere. The hosts it links most: docs.perplexity.ai (183), community.perplexity.ai (1), research.perplexity.ai (1).
| H2 section | Links |
|---|---|
| OpenAPI Specs | 5 |
| Optional | 2 |
Grade against the llmstxt.org structure
The file passes all six checks. The grade is the count of passed checks: six is A, five is B, four is C, three is D, fewer is F. It measures structure only. Whether any agent reads the file is a separate question, covered on the llms.txt reference.
| Check | Result | Detail |
|---|---|---|
| Starts with an H1 naming the site or project | Pass | H1: Perplexity |
| Has a blockquote summary under the H1 | Pass | Blockquote summary found. |
| Has at least one H2 section that lists links | Pass | 2 H2 sections, 2 with links. |
| At least 80 percent of links point at the same site | Pass | 190 of 190 links stay on perplexity.ai (100 percent). |
| Under 100 KB | Pass | 42749 bytes as served. |
| Plain markdown, no HTML | Pass | No HTML document markers in the body. |
Ora's cached score
Ora's cached report for docs.perplexity.ai gives it 76 of 100, grade B, scanned on 24 August 2026. Its Access layer, which holds the read-layer checks below, scored 55 of 64 check points. Read from GET ora.ai/api/score/docs.perplexity.ai on 5 November 2026. The score is Ora's, not this site's, and the structure grade above is a different measure: it grades the llms.txt file against the llmstxt.org proposal, while Ora's score covers the whole site. Ora's report for docs.perplexity.ai has every layer.
| Ora check (Access layer) | Result | Points |
|---|---|---|
| Sitemap exists | Pass | 2 of 2 |
| Content without JavaScript | Partial | 1 of 3 |
| Not blocked by bot detection | Pass | 2 of 2 |
| Markdown URL fallback | Pass | 2 of 2 |
| llms.txt exists | Pass | 1 of 1 |
| llms.txt formatting | Partial | 1 of 2 |
| JSON-LD structured data | Pass | 3 of 4 |
| Metadata completeness | Pass | 2 of 2 |
| llms.txt links resolve | Pass | 2 of 2 |
| Markdown alternate link | Pass | 1 of 1 |
| Redirect hygiene | Pass | 1 of 1 |
| Page token budget | Pass | 1 of 1 |
| Code fence validity | Pass | 1 of 1 |
| Content behind auth | Pass | 2 of 2 |
| OpenAPI spec published | Pass | 7 of 7 |
| Markdown agent docs | Pass | 1 of 1 |
| Markdown negotiation with Vary | Partial | 0 of 1 |
| Markdown for agent user agents | Fail | 0 of 1 |
| Agent crawler reachability | Pass | 2 of 2 |
Linking and the companion file
The home page of docs.perplexity.ai links the file in its HTML. Ora's lab found (2026-10-19) that 86 percent of llms.txt fetches by agents arrived through a link, so an unlinked file is mostly a file nobody opens.
How it compares
At 42 KB the file is larger than 90 percent of the 2,529 plain-text llms.txt files in this edition, and its 190 links put it ahead of 89 percent of them on link count. The home page declares no generator. Size bucket: 10 to 50 KB. Curated example. Category: AI lab.
Why it is in the index: Answer-engine operator publishing its own API docs index. Owner as listed: Perplexity.
Live file: https://docs.perplexity.ai/llms.txt. Read 4 November 2026. If the file has since changed or gone, the live URL is the authority; this page records what the crawl read on that date.
Sources
- docs.perplexity.ai llms.txt https://docs.perplexity.ai/llms.txt, read 4 November 2026
- llms.txt proposal (llmstxt.org) https://llmstxt.org/, read 4 November 2026
- Tranco list 746WX (crawl universe) https://tranco-list.eu/list/746WX, read 4 November 2026
- Ora: AgentReady launch post (the 86 percent finding) https://ora.ai/blog/agentready-launch, read 4 November 2026
- Ora llms.txt (GET /api/score/{domain} looks up a cached score) https://ora.ai/llms.txt, read 5 November 2026
Last reviewed 5 November 2026