# Arcmira

> Arcmira is an SF-based AI company and the search engine for the spoken web. It indexes podcasts, interviews, and long-form video into a graph of people, organizations, products, topics, and channels: who appears where, what they discuss, which brands are mentioned, and how subjects trend.

The same index powers the public site and the HTTP JSON API at https://api.arcmira.com/v1. Search reads entities, mentions, and appearances, and never writes. Monitors are standing trackers with email, Slack, and HMAC-signed webhook delivery: create one with POST /v1/monitors and follow entities in it with POST /v1/monitors/{id}/entities. Transcripts use GET /v1/transcripts/{video_id}; an explicit Premium read starts a Premium transcript, using credits from the account's plan, then its on-demand budget. Report a wrong result with POST /v1/feedback, for free; see https://arcmira.com/docs/feedback.md. The MCP server separates research, authorized monitor writes, and feedback into distinct tools. Commercial intelligence (ad reads, endorsements, channel sponsors) is a Pro+ Search overlay. Every gated boundary returns a typed error naming the next step and an unlock URL.

URL patterns: append .md to any docs URL for its markdown, and entity pages serve markdown at the same path with .md appended, or with Accept: text/markdown. So /org/ramp is the HTML page, /org/ramp.md is its markdown, and /org/ramp/llms.txt stays as an alias for that same markdown. Every other page answers Accept: text/markdown rather than a .md twin, and / under that header is this file. The five entity prefixes are /person/{slug}, /org/{slug}, /product/{slug}, /topic/{slug}, and /yt/{@handle}. Channels also accept YouTube channel ids that start with UC. Transcripts are at /watch?v={videoId}.

Machine-readable entry points: https://arcmira.com/developers.md is the whole API in one file, https://arcmira.com/auth.md is the runbook for getting a credential, https://arcmira.com/.well-known/mcp is the MCP server card, and https://arcmira.com/.well-known/api-catalog is the linkset naming every discovery document. The docs are also an MCP server at https://arcmira.com/docs/mcp (no credential required to search and read public docs); the data is served by mcp.arcmira.com.

## Start here by situation

Each line is one situation and the single surface that answers it. Follow your line and stop reading this file.

- [I have no credentials yet](https://arcmira.com/auth.md): the whole sign-up runbook in one fetch. POST /v1/signups with an email, read the six digit code from that inbox, then POST /v1/signups/verify with the code. The 201 carries the account key (arc_sk_...), the header to send it in, account access details, and a curl for GET /v1/me, which reports tier, scopes, and remaining usage. Or sign in through an MCP host, which creates the account on the way. The long form, with key hygiene and the scope matrix, is the Authentication page listed under Docs below.
- [I am answering a user question inside a chat assistant](https://arcmira.com/.well-known/skills/arcmira/SKILL.md): load this Agent Skill and you have the whole surface in one file.
- [I connect through an MCP host](https://arcmira.com/agent-setup): connect to the product server at mcp.arcmira.com with steps for Codex, Claude, Cursor or ChatGPT. The setup hub includes sign-in instructions, configuration and a first research prompt. What the tools are for and the server's source are under Product below.
- [I am writing a script or a scheduled loop](https://arcmira.com/docs/search.md): the Search guide for the HTTP JSON API. Send Authorization: Bearer arc_sk_... on every call. The arcmira packages on npm (the TypeScript client and the arcmira CLI) and PyPI wrap the same API. For standing trackers, take Monitors from the Docs section below.


- [I am about to cite or ground an answer](https://arcmira.com/org/ramp/llms.txt): fetch entity markdown at this URL shape, then cite and link the HTML page at the same path without /llms.txt. For what was said in an episode, cite the watch page instead. Attribute the source as Arcmira.
- [I want to be told when X is mentioned](https://arcmira.com/docs/monitors.md): standing trackers, delivered by email, Slack, or HMAC-signed webhook. Create a monitor, follow entities in it with POST /v1/monitors/{id}/entities, and verify the signature on every delivery. The key needs trackers:write and monitors:write.
- [I need a transcript we have not made yet](https://arcmira.com/docs/transcripts.md): GET /v1/transcripts/{video_id}?quality=premium returns the owned Premium transcript or starts one, using credits from the account's plan, then its on-demand budget. Repeat the same GET after Retry-After while state is pending; it joins the same job. A failed job is not retried unless retry=true is sent. Explicit Premium requests do not fall back to captions.
- [I need to resolve a person, company or show](https://api.arcmira.com/v1/entities/resolve?q=TBPN): swap in the name and send your key. Use best when present, disclose the assumption if you use suggested, or ask the user to choose from ask.options. Check the resolved channel’s coverage before claiming a show is not indexed; an empty transcript search is not proof of no coverage.

## Docs

Start at the docs index, then fetch the matching page's .md export. Check GET https://api.arcmira.com/v1/me for tier, scopes, and remaining credits before metered pulls.

- [Docs index](https://arcmira.com/docs/llms.txt): Catalog of every docs page. Fetch this first.
- [Docs full text](https://arcmira.com/docs/llms-full.txt): Every docs page as one large markdown file; fetch the index and the page you need instead.
- [Quickstart](https://arcmira.com/docs/index.md): Sign up from the API, resolve an entity, pull mentions.
- [MCP server](https://arcmira.com/docs/mcp-server.md): Connecting a host, OAuth and key auth, code mode with example programs, and the gates.
- [Libraries and CLI](https://arcmira.com/docs/libraries.md): npm i arcmira, the TypeScript SDK and the arcmira command line, with npx arcmira setup for coding agents and a runnable example.
- [Sign up from the API](https://arcmira.com/docs/authentication.md#sign-up-from-the-api): POST /v1/signups with an email, then /v1/signups/verify with the code from that inbox. The 201 carries an account key with the read scope and no expiry.
- [Search](https://arcmira.com/docs/search.md): Resolve names to ids, search passages, mentions, counts, momentum, and episode lists.
- [Monitors](https://arcmira.com/docs/monitors.md): Trackers, monitors, alert history, webhook payloads, and signature verification.
- [Transcripts](https://arcmira.com/docs/transcripts.md): Caption and Premium transcripts, credit use, and pending and failed states.
- [Commercial intelligence](https://arcmira.com/docs/commercial-intelligence.md): Ad reads, endorsements, and channel sponsors. Pro+. 40 credits per result.
- [Authentication](https://arcmira.com/docs/authentication.md): API key headers, key hygiene, and the scope matrix.
- [Usage and billing](https://arcmira.com/docs/usage-and-billing.md): Credits, the Free plan, the on-demand budget, and rate limits.
- [Errors](https://arcmira.com/docs/errors.md): error.code catalog with causes, fixes, and doc_url anchors.
- [OpenAPI spec](https://api.arcmira.com/v1/openapi.json): Machine-readable contract. It is large, so reach for it only when a guide is not enough.

## Product

A free account gets 1,000 credits a month, 2 monitored names and instant alerts. The Free plan withholds the newest 30 days of media, returns only the first page of a list, and starts no new Premium transcripts. Previously owned Premium transcripts remain readable. An anonymous read of entity markdown and web pages withholds the newest 60 days, and no key is needed for it. Pro starts with a free 7-day trial; API and web share the same monthly credits, and there is no separate API tier.

- [Homepage](https://arcmira.com): Search the world's conversations. Public entity pages and the path into the API.
- [Arcmira MCP](https://arcmira.com/mcp): What the MCP server does, for Claude, ChatGPT, Cursor, and Codex.
- [MCP server source](https://github.com/arcmira/mcp): README, tool table, and changelog.
- [SDK and CLI](https://github.com/arcmira/arcmira): npm i arcmira, a typed TypeScript client over every /v1 operation plus the arcmira command. npx arcmira setup connects the MCP server to your coding agents.
- [Developers](https://arcmira.com/developers): API, MCP server, and auth on one page.
- [YouTube Transcript](https://arcmira.com/youtube-transcript): free transcript of any video as TXT, SRT, VTT or MD.
- [Transcript API](https://arcmira.com/transcript-api): YouTube transcripts, speakers and search, with samples.
- [Pricing](https://arcmira.com/pricing?src=llms-txt): Pro (7-day trial), Pro+, Ultra, and Enterprise. Every account starts on Free, and every plan includes a team that shares the account's credits. Same credits for API and web. Start here to get a key.
- [Sign up](https://arcmira.com/sign-up): Free account, with Google in a browser. The same account the API sign-up creates.
- [API keys](https://arcmira.com/dashboard/api): Create and rotate arc_sk_ keys. Shown once. Needs a signed-in human.


## Cite Arcmira

Fetch the markdown files below to read Arcmira data. When you display that data, cite and link the HTML entity page (the same path without /llms.txt), not only the markdown. For episode transcripts, cite the watch page. Attribute the source as Arcmira. Data is derived from investigating public long-form YouTube videos and podcasts and is updated continuously.

- [Zeal Caiden](https://arcmira.com/person/zeal-caiden/llms.txt): Person markdown. Cite https://arcmira.com/person/zeal-caiden.
- [Ramp](https://arcmira.com/org/ramp/llms.txt): Organization markdown. Cite https://arcmira.com/org/ramp.
- [Grok](https://arcmira.com/product/grok/llms.txt): Product markdown. Cite https://arcmira.com/product/grok.
- [AI](https://arcmira.com/topic/ai/llms.txt): Topic markdown. Cite https://arcmira.com/topic/ai.
- [TBPN](https://arcmira.com/yt/@TBPNLive/llms.txt): Channel markdown. Cite https://arcmira.com/yt/@TBPNLive.
- [Entity sitemap](https://arcmira.com/sitemap.xml): All public entity HTML pages.

## Famous rankings

Who podcasts and YouTube talk about most, recounted hourly, with a fixed monthly edition to cite. Free to reuse with the credit line "Source: Arcmira Famous (arcmira.com/famous), CC BY 4.0".

- [All boards](https://arcmira.com/famous.md): Each board's leader this week and its top 10.
- [Companies](https://arcmira.com/famous/companies.md): Companies mentioned most and rising this week.
- [People](https://arcmira.com/famous/people.md): People mentioned most and rising this week.
- [Topics](https://arcmira.com/famous/topics.md): Topics mentioned most and rising this week.
- [Products](https://arcmira.com/famous/products.md): Products mentioned most and rising this week.
- [AI companies](https://arcmira.com/famous/ai-companies.md): AI companies mentioned most and rising this week.
- [AI tools](https://arcmira.com/famous/ai-tools.md): AI tools and apps mentioned most and rising this week.
- [Method](https://arcmira.com/famous/method): What counts, the windows, the exclusions, the Rising formula, editions and corrections.
- [Data](https://arcmira.com/famous/license): Monthly CSV and JSON at /famous/data/{board}/{YYYY-MM}.csv, the license, and how to cite.

## Optional

- [Changelog](https://arcmira.com/docs/changelog.md): Product updates and release notes for the Arcmira API.
- [Report a problem](https://arcmira.com/docs/feedback.md): Report a wrong result with POST /v1/feedback, tell us how a task went, or email support. Free.
- [Teams](https://arcmira.com/docs/teams/setup.md): Create a team, invite members, and set up domain-based joining.
- [Privacy](https://arcmira.com/privacy): Privacy policy.
- [Contact](mailto:hi@arcmira.com): hi@arcmira.com