> ## Documentation Index > Fetch the complete documentation index at: https://docs.piramyd.cloud/llms.txt > Use this file to discover all available pages before exploring further. # POST /v1/llmstxt — Generate llms.txt for Any Website > POST /v1/llmstxt — generate an llms.txt file for any website: a structured plain-text document of the site's content optimized for LLM context injection. An [`llms.txt`](https://llmstxt.org/) file is a structured plain-text document that summarizes a website's content in a format optimized for injection into an LLM's context window. Instead of scraping dozens of pages individually and concatenating their markdown, `POST /v1/llmstxt` does that work for you — it crawls the site, extracts the most relevant content from up to `max_urls` pages, and compiles everything into a single well-structured document. AI agents can load this document at the start of a session to ground the model in a site's full knowledge base with a single API call. Set `show_full_text: true` to also generate an `llms-full.txt` variant with complete, untruncated page content. ## Endpoint ``` POST https://api.piramyd.cloud/v1/llmstxt ``` **Authentication:** `Authorization: Bearer sk-` *** ## Request Parameters The root URL of the website to generate `llms.txt` for (e.g. `https://fastapi.tiangolo.com`). Maximum number of pages to include in the generated document. Higher values produce more comprehensive output but take longer and consume more credits. Accepted range: **1–100**. When `true`, the response also includes `llmsfulltxt` — a second document containing the complete, untruncated content of every included page. Useful when you need exhaustive context rather than a summary-optimized document. Maximum time in seconds for the generation job to complete synchronously. Accepted range: **10–120**. Larger sites may return `status: "processing"` with a `job_id` to poll. *** ## Response `true` if the generation request was accepted without error. Unique identifier for this generation job. Present on all responses — use it to poll when `status` is `"processing"`. Generation status. Either `"completed"` (content is in this response) or `"processing"` (the job is still running — poll with `job_id`). The generated `llms.txt` content — a structured, summary-optimized plain-text document of the site. Present when `status` is `"completed"`. The generated `llms-full.txt` content with complete page text. **Only present when `show_full_text: true` and `status` is `"completed"`.** *** ## Code Example ```python Python theme={null} import httpx import time PIRAMYD_KEY = "sk-YOUR_KEY" BASE = "https://api.piramyd.cloud/v1" HEADERS = {"Authorization": f"Bearer {PIRAMYD_KEY}"} resp = httpx.post( f"{BASE}/llmstxt", headers=HEADERS, json={ "url": "https://docs.piramyd.cloud", "max_urls": 30, "show_full_text": False, "timeout_seconds": 90, }, timeout=100, ) resp.raise_for_status() result = resp.json() # Handle async case if result["status"] == "processing": job_id = result["job_id"] print(f"Generation queued as {job_id}, polling...") while result["status"] == "processing": time.sleep(5) poll = httpx.get(f"{BASE}/llmstxt/{job_id}", headers=HEADERS) result = poll.json() # Use the generated content llmstxt_content = result["llmstxt"] print(f"Generated {len(llmstxt_content)} characters of llms.txt") print("\n--- llms.txt preview ---") print(llmstxt_content[:800]) ``` ```python Python (with full text) theme={null} import httpx PIRAMYD_KEY = "sk-YOUR_KEY" BASE = "https://api.piramyd.cloud/v1" HEADERS = {"Authorization": f"Bearer {PIRAMYD_KEY}"} resp = httpx.post( f"{BASE}/llmstxt", headers=HEADERS, json={ "url": "https://fastapi.tiangolo.com", "max_urls": 20, "show_full_text": True, "timeout_seconds": 120, }, timeout=130, ) resp.raise_for_status() result = resp.json() # Save both documents with open("llms.txt", "w") as f: f.write(result["llmstxt"]) if result.get("llmsfulltxt"): with open("llms-full.txt", "w") as f: f.write(result["llmsfulltxt"]) print("Saved llms.txt and llms-full.txt") ``` ```json Example request body theme={null} { "url": "https://fastapi.tiangolo.com", "max_urls": 20, "show_full_text": false, "timeout_seconds": 60 } ``` ```json Example response theme={null} { "success": true, "job_id": "llmstxt_abc456", "status": "completed", "llmstxt": "# FastAPI\n\n> Modern, fast web framework for building APIs with Python\n\n## Docs\n\n- [Tutorial](https://fastapi.tiangolo.com/tutorial/): Step-by-step guide to building your first API...\n- [Advanced](https://fastapi.tiangolo.com/advanced/): Advanced user guide topics...", "llmsfulltxt": null } ``` *** ## Errors | Error code | HTTP status | Description | | ------------------------ | ----------- | ------------------------------------------------------------------------------------------------------------ | | `llmstxt_timeout` | 504 | Generation did not complete within `timeout_seconds`. Increase the timeout or reduce `max_urls`, then retry. | | `llmstxt_upstream_error` | 502 | CROWD returned an upstream error during content extraction. Retry with backoff. | | `llmstxt_unavailable` | 502 | The llms.txt generation service is temporarily unavailable. Retry with backoff. | | `llmstxt_internal_error` | 500 | Unexpected internal error. Retry with backoff. | For large sites (`max_urls` > 20), set `timeout_seconds` to `120` and your HTTP client timeout to at least 130 seconds. Jobs that exceed the synchronous timeout return `status: "processing"` — poll the `job_id` until `status` is `"completed"`. You can use `POST /v1/llmstxt` on Piramyd's own docs to keep your agent up to date with the latest API capabilities — just point `url` at `https://docs.piramyd.cloud` and inject the result into your system prompt.