Skip to content
J·BJon BishopFractional growth marketer

Artificial intelligence · llms.txt file breakdown

Vapi llms.txt

A directly retrieved Markdown index for Vapi's documentation, with 449 links across 5 detected headings.

The actual file

Source Markdown

The saved llms.txt text, displayed as source. The file belongs to its publisher; the live endpoint may change.

Open saved text ↗

Showing the first 18000 characters. Load the complete file to read the remainder.

# Vapi

## Instructions for AI Agents

- For clean Markdown of any page, append `.md` to the page URL
- For section-specific indexes, append `/llms.txt` to any section URL
- For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.vapi.ai/_mcp/server

## Docs

- [Introduction](https://docs.vapi.ai/quickstart/introduction.md): Get started with Vapi, the developer platform for building voice AI agents that make and receive phone calls, handling the infrastructure so you can build.
- [Phone calls](https://docs.vapi.ai/quickstart/phone.md): Make your first Vapi phone call in minutes, creating a voice agent in the Dashboard or programmatically and talking to it over the phone.
- [Web calls](https://docs.vapi.ai/quickstart/web.md): Build voice interfaces and backend integrations with Vapi's Web and Server SDKs, covering client-side voice apps and server-side call management.
- [Guides](https://docs.vapi.ai/guides.md): Browse real-world, cloneable Vapi examples that build voice agents with assistants and squads, from appointment booking to multilingual support routing.
- [Composer](https://docs.vapi.ai/composer.md): Build and configure Vapi voice agents through natural conversation with Composer, which creates assistants, sets up numbers and integrations, and troubleshoots.
- [Vapi CLI](https://docs.vapi.ai/cli.md): Command-line interface for building voice AI applications faster
- [Agent skills](https://docs.vapi.ai/agent-skills.md): Learn how to install agent skills for Vapi and use them with AI coding assistants to create assistants, configure tools, place calls, and run Simulations.
- [Config as code (GitOps)](https://docs.vapi.ai/config-as-code.md): Use Vapi GitOps to manage assistants, squads, tools, structured outputs, simulations, and evals as code, then review and promote changes across environments.
- [Path to production](https://docs.vapi.ai/quickstart/path-to-production.md): Take a Vapi assistant from a first test call to production. Choose a model preset, publish a version, test it with Simulations, and monitor live calls.
- [Assistants quickstart](https://docs.vapi.ai/assistants/quickstart.md): Build your first Vapi assistant, connect it to a phone number, and make a call, working through the steps in the Dashboard or with the Vapi API.
- [Transient vs permanent configurations](https://docs.vapi.ai/assistants/concepts/transient-vs-permanent-configurations.md): Choose between transient inline and permanent stored assistant configurations in Vapi, and learn when each fits for flexibility, reusability, and management.
- [Variables](https://docs.vapi.ai/assistants/dynamic-variables.md): Personalize Vapi assistant prompts and messages with dynamic variables in double curly braces, setting values at call time through assistantOverrides.
- [Multilingual support](https://docs.vapi.ai/customization/multilingual.md): Configure multilingual voice AI agents with automatic language detection, cross-language conversation, and localized voices
- [Personalization with user information](https://docs.vapi.ai/assistants/personalization.md): Personalize Vapi conversations with customer-specific data your server supplies at call time, tailoring responses for support, account management, and more.
- [Voice formatting plan](https://docs.vapi.ai/assistants/voice-formatting-plan.md): Format Vapi LLM output for natural-sounding speech, expanding numbers, currency, and symbols so text-to-speech reads them clearly, enabled by default.
- [Flush syntax](https://docs.vapi.ai/assistants/flush-syntax.md): Force immediate voice transmission with VAPI's flush syntax for real-time interactions
- [Background messages](https://docs.vapi.ai/assistants/background-messages.md): Silently add information to a Vapi call's chat history without notifying the user, logging button presses, actions, or context updates during the conversation.
- [Idle messages](https://docs.vapi.ai/assistants/idle-messages.md): Keep users engaged during Vapi call pauses with idle messages that trigger on customer silence, using fixed text or model-generated prompts via assistant hooks.
- [Assistant hooks](https://docs.vapi.ai/assistants/assistant-hooks.md): Automate actions on Vapi call events like ending and speech interruptions, using assistant hooks to transfer calls, run functions, or send messages.
- [Call timeout settings](https://docs.vapi.ai/documentation/assistants/conversation-behavior/call-timeout-settings.md): Configure silence and maximum-duration timeouts for assistants, distinguish inactivity from total call duration, and troubleshoot timeout call endings.
- [Keypad input settings](https://docs.vapi.ai/documentation/assistants/conversation-behavior/keypad-input-settings.md): Keypad input settings let assistants collect caller-entered digits. Configure timeouts and delimiters for menus, codes, and account information during calls.
- [End call phrases](https://docs.vapi.ai/documentation/assistants/conversation-behavior/end-call-phrases.md): End call phrases let assistants end calls after saying configured text. Choose reliable phrases and prevent unintended transcript matches during calls.
- [Background speech denoising](https://docs.vapi.ai/documentation/assistants/conversation-behavior/background-speech-denoising.md): Filter out noise and background speech while users are talking
- [Pronunciation dictionaries](https://docs.vapi.ai/assistants/pronunciation-dictionaries.md): Control how a Vapi assistant pronounces specific words, names, and acronyms using pronunciation dictionaries with phoneme and alias rules by voice provider.
- [Email address reading](https://docs.vapi.ai/assistants/email-address-reading.md): Get a Vapi voice agent to collect, read back, and confirm email addresses clearly, using built-in TTS formatting, prompting, and confirmation techniques.
- [Speech configuration](https://docs.vapi.ai/customization/speech-configuration.md): Control when a Vapi assistant starts and stops speaking, tuning speech settings to sound natural, avoid interrupting the customer, and reduce awkward pauses.
- [Voice pipeline configuration](https://docs.vapi.ai/customization/voice-pipeline-configuration.md): Complete guide to configuring VAPI's voice pipeline for optimal conversation timing and interruption handling
- [Voice fallback configuration](https://docs.vapi.ai/voice-fallback-plan.md): Configure an ordered voice fallback plan in the Dashboard or API, test backup voices, and troubleshoot calls that end before fallback activates.
- [Transcriber fallback configuration](https://docs.vapi.ai/customization/transcriber-fallback-plan.md): Configure fallback transcribers so calls keep working if your primary speech-to-text provider fails. Covers automatic fallback, manual fallback, and HIPAA-compliant provider options.
- [Model fallback configuration](https://docs.vapi.ai/customization/model-fallback-plan.md): Configure fallbackModels for OpenAI and Anthropic Bedrock models in the Dashboard or API, and compare model fallback with voice and transcriber fallback.
- [OpenAI Realtime](https://docs.vapi.ai/openai-realtime.md): Build Vapi voice assistants on OpenAI's Realtime speech-to-speech models for low latency, covering model choice, configuration, and production best practices.
- [Provider Keys](https://docs.vapi.ai/customization/provider-keys.md): Connect provider credentials in Vapi and find configuration details for transcriber, model, voice, and cloud providers.
- [Versioning](https://docs.vapi.ai/assistants/versioning.md): Versioning saves published assistant and tool configurations so you can make changes in a draft, review their history, and restore an earlier version.
- [Versioning with assistants](https://docs.vapi.ai/assistants/versioning/versioning-assistants.md): Publish assistant changes, review version history, restore an earlier configuration as a new current version, and select the published assistant version for a call.
- [Versioning with tools](https://docs.vapi.ai/assistants/versioning/versioning-tools.md): Publish tool changes, review version history, restore an earlier configuration to the draft, and select the specific tool version that an assistant uses.
- [How versioning works with assistants and tools](https://docs.vapi.ai/assistants/versioning/versioning-with-assistants-and-tools.md): Select a specific or latest tool version for an assistant, and learn how restoring an assistant also restores the tool version selections saved with it.
- [How squad versioning works with assistant versions](https://docs.vapi.ai/assistants/versioning/versioning-with-squads.md): Create and review squad versions, pin members to specific published assistant versions or use Latest, and choose a squad version for outbound calls.
- [Traffic splitting](https://docs.vapi.ai/assistants/versioning/traffic-splitting.md): Split live call traffic between published assistant versions by percentage, so you can canary a new version, run tests, and roll back instantly.
- [Model Intelligence for transcribers, models, and voices](https://docs.vapi.ai/assistants/model-intelligence/overview.md): Model Intelligence helps you choose transcribers, models, and voices with curated presets and weekly refreshed performance metrics for latency, cost, and quality.
- [Model Presets for transcribers, models, and voices](https://docs.vapi.ai/assistants/model-intelligence/presets.md): Model presets optimize for balanced performance, reasoning, low latency, or low cost. Learn when to use each preset and how to apply or customize its configuration.
- [Performance metrics and methodology reference](https://docs.vapi.ai/assistants/model-intelligence/metrics-methodology.md): The performance metrics reference explains how Vapi sources and calculates latency, cost, and quality data for each supported transcriber, model, and voice.
- [How voice agent component costs and estimates work](https://docs.vapi.ai/assistants/model-intelligence/understanding-cost.md): Understand how Vapi estimates transcriber, model, and voice costs per minute, including prompt caching assumptions and the effects of conversation history.
- [How voice-to-voice latency works in voice agents](https://docs.vapi.ai/assistants/model-intelligence/understanding-latency.md): Understand voice-to-voice latency, the components included in Vapi performance metrics, and how model capability, endpointing, and transport affect response time.
- [Tools](https://docs.vapi.ai/tools.md): Learn what a Vapi tool is, how a tool call flows between the model, Vapi, and your server, which tool types exist, and the configuration every tool shares.
- [Built-in tools](https://docs.vapi.ai/tools/default-tools.md): Add Vapi's built-in default tools to assistants, including transfer call, hang up, SMS, DTMF keypad, and API request, through the API or dashboard.
- [When to use API Request or Function tools](https://docs.vapi.ai/tools/api-request-vs-function.md): Compare Vapi's API Request and Function tools, choosing API Request for direct HTTPS APIs and Function for tool-calls webhooks or async and client-side execution.
- [API Request tool](https://docs.vapi.ai/tools/api-request.md): Learn when to use an API Request Tool, how requests move between an assistant and an HTTP endpoint, and how to build your first working integration with Vapi.
- [API Request tool quickstart](https://docs.vapi.ai/tools/api-request/quickstart.md): Build a coffee-order assistant that sends confirmed orders to a public endpoint and reads the server-calculated total and order number back to the caller.
- [Customize API Request tool requests](https://docs.vapi.ai/tools/api-request/configuration.md): Configure an API Request Tool's URL, method, model-generated body, request headers, static body fields, and Liquid values with Dashboard and cURL examples.
- [Get order status](https://docs.vapi.ai/tools/api-request/order-status.md): Add an order-status API Request Tool to the coffee-order assistant, update its opening and system prompt, and verify status lookup across two inbound calls.
- [Handle API Request Tool latency and retries](https://docs.vapi.ai/tools/api-request/reliability.md): Learn how API Request Tool timeouts and failed HTTP requests behave, keep callers informed, and handle safe retries in a service you control.
- [Handle API Request tool responses and errors](https://docs.vapi.ai/tools/api-request/response-handling.md): Handle successful and failed API Request Tool responses, test structured errors, guide assistant recovery, and inspect call results in the Dashboard or with cURL.
- [Use API Request tool response data](https://docs.vapi.ai/tools/api-request/response-data.md): Use API Request Tool JSON responses immediately, extract response fields with schemas or aliases, and pass extracted variables to subsequent tools.
- [Function tools](https://docs.vapi.ai/tools/custom-tools.md): Create and manage Function tools through the Vapi dashboard or API, connect them to server-side logic, test webhooks locally, and reuse them across assistants.
- [Transfer Call tool](https://docs.vapi.ai/tools/transfer-call.md): Create a transfer call tool in the Vapi Dashboard or API, add it to an assistant, configure destinations, and troubleshoot invocation and routing failures.
- [Configure blind transfer modes](https://docs.vapi.ai/tools/transfer-call/blind-transfer.md): Configure blind call transfers in the Vapi Dashboard or API, choose a SIP transfer method, add a summary header, verify routing, and diagnose failures.
- [Configure warm transfer modes](https://docs.vapi.ai/tools/transfer-call/warm-transfer.md): Configure traditional warm transfer modes in the Vapi Dashboard or API, including messages, summaries, operator-first timing, TwiML, and timeouts.
- [Configure assistant-based warm transfer](https://docs.vapi.ai/calls/assistant-based-warm-transfer.md): Configure an assistant-based warm transfer in the Vapi Dashboard or API, including operator prompts, fallback behavior, summaries, custom audio, and testing.
- [Hang Up tool](https://docs.vapi.ai/tools/hang-up.md): Learn what the Hang Up tool does, when to use it instead of end call phrases, and how an assistant ends a call with a configured goodbye message.
- [Hang Up tool quickstart](https://docs.vapi.ai/tools/hang-up/quickstart.md): Build an assistant for the Vapi hardware store that answers a caller's question, confirms they have nothing else to ask, says goodbye, and ends the call with the Hang Up tool.
- [Configure the Hang Up tool](https://docs.vapi.ai/tools/hang-up/configuration.md): Configure the Hang Up tool's goodbye message, attach it to an assistant, write prompt rules, add a rejection plan, and confirm how calls end with Dashboard and cURL examples.
- [Handoff tool](https://docs.vapi.ai/squads/handoff.md): Transfer a Vapi call between assistants with the handoff tool, covering destination types, context management, and configuration for multi-agent squads.
- [Configure handoff destinations](https://docs.vapi.ai/squads/handoff/destinations.md): Declare handoff destinations by assistant ID or name, offer the model several options, resolve one at runtime by webhook, or hand off to a whole squad.
- [Pass context and variables](https://docs.vapi.ai/squads/handoff/context-and-variables.md): Choose the conversation history a destination receives and pass variables through handoff arguments, tool results, or extraction.
- [Configure messages and rejection rules](https://docs.vapi.ai/squads/handoff/messages-and-rejection.md): Set the messages an assistant speaks at each phase of a handoff, and block handoff attempts using regex, Liquid, or grouped conditions on conversation state.
- [Troubleshoot handoffs](https://docs.vapi.ai/squads/handoff/troubleshooting.md): Fix handoffs that never fire or route to the wrong assistant, starting with the system prompt instructions that tell the model when to call the handoff tool.
- [Voicemail tool](https://docs.vapi.ai/tools/voicemail-tool.md): Give a Vapi assistant direct control over voicemail with the voicemail tool, deciding when it reaches voicemail to leave a message or end the call.
- [Configure the Voicemail tool](https://docs.vapi.ai/tools/voicemail-tool/configure.md): Create a Vapi voicemail tool in the Dashboard or API, attach it to an assistant, prompt for voicemail detection, and set the message for each behavior mode.
- [Client-side tools (Web SDK)](https://docs.vapi.ai/tools/client-side-websdk.md): Handle Vapi tool calls entirely in the browser with the Web SDK, triggering UI side effects like notifications or state changes without a server URL.
- [Knowledge bases](https://docs.vapi.ai/knowledge-base.md): Create and integrate knowledge bases into Vapi voice agents so they retrieve information from your documents to answer questions accurately during a call.
- [Query tool](https://docs.vapi.ai/knowledge-base/using-query-tool.md): Configure the Vapi query tool to give an assistant access to a knowledge base, retrieving information from your files by ID for accurate, grounded responses.
- [Custom knowledge base](https://docs.vapi.ai/knowledge-base/custom-knowledge-base.md): Build a custom knowledge base in Vapi with your own retrieval server, using your vector database or search logic to control how an assistant finds answers.
- [Implement the custom endpoint](https://docs.vapi.ai/knowledge-base/implement-custom-endpoint.md): Handle the Vapi knowledge base request on your own server, return documents or a direct answer, integrate a vector database, and harden the endpoint.
- [Integrations](https://docs.vapi.ai/tools/integrations.md): Connect a Vapi assistant to external services using prebuilt integrations or service guides for API Request, Model Context Protocol (MCP), and Function tools.
- [Choose a Make integration pattern](https://docs.vapi.ai/tools/integrations/make.md): Choose a Make integration pattern for an assistant: API Request for webhooks, Model Context Protocol (MC

Measured from the saved file

Linked document

UTF-8 text size
85.3 KB87391 bytes
Token count
19221o200k_base encoding
Markdown links
453453 distinct URLs
Fenced code blocks
0470 lines in the file

Size measures the captured UTF-8 text, not compressed transfer size. Tokens are counted with a named tokenizer; other models can count differently. Neither measure establishes context compatibility or answer quality.

Project H1
Vapi

Context architecture

How the content is divided

4 H2 sections organize 453 parsed Markdown links. 450 links point to .md or .markdown paths; 0 link occurrences use another hostname. Linked destinations have not been recursively validated.

H2 sectionLinksUTF-8 bytesTokens
Instructions for AI AgentsLine 3025465
DocsLine 93086846614971
API DocsLine 320143184194118
OpenAPI SpecificationLine 466216949

Nested indexes

1 distinct links point to other llms.txt paths. These are routes to more indexes, not proof their contents were included in this file.

Link counts exclude parsed fenced and inline code and resolve relative destinations against the final endpoint. Section sizes cover the text below each H2 until the next H2.

Embedded material

Code, installation and API examples

Fenced blocks
0 closed
Declared languages
No fenced blocks detected
Package-install commands
0 detected in fenced or inline code
Authentication examples
0 fenced lines match API-key, Authorization, Bearer, OAuth or access-token patterns
HTTP request examples
0 fenced lines match request-method patterns
Authentication links
0 links have authentication-related labels

Embedded example text and links to further documentation are counted separately. A code block does not establish that a sample runs or that an authentication flow succeeds.

Publisher prioritization

The Optional section

No H2 section named “Optional” was detected in this file.

The community proposal uses Optional by convention for secondary links that can be skipped for a shorter context. This label does not establish that everything outside it is critical, or that the material inside it is deprecated.

Machine-readable routes

MCP references and configuration text

MCP-related text was detected in the captured file. 3 links have MCP-related labels or paths. 0 fenced lines match mcpServers, mcp_servers or an mcp add command.

These are text and link observations. No MCP server connection, authentication or tool call was tested.

Index and bundle

llms-full.txt companion

Verified text companion

Text companion verified

https://docs.vapi.ai/llms-full.txt ↗
Discovery
Conventional sibling path checked
HTTP response
206 · text/plain; charset=utf-8
Text size
85.3 KB · 87391 bytes (server-reported)
Token count
19221 o200k_base tokens
Retrieval scope
Complete body retrieved

A verified text response establishes a companion at the checked URL. It does not prove the bundle contains all product documentation, that generation is automated, or that it fits an assistant’s context window.

Change evidence

What can be said about freshness

This capture establishes a baseline. Update frequency cannot be determined from one observation.

Linked release and migration material

2 links have release, changelog, deprecation or migration terms in their labels or paths. These provide places to investigate product changes separately.

File changes can reveal documentation edits or navigation changes. They do not by themselves measure product release velocity, accuracy, agent adoption or AI visibility.

Capture identity and measurement method
Requested endpoint
https://docs.vapi.ai/llms.txt
Final endpoint
https://docs.vapi.ai/llms.txt
HTTP status
200
Content type
text/plain; charset=utf-8
Text retrieval
Firecrawl literal text capture
Tokenizer
tiktoken 0.14.0 · o200k_base
SHA-256 of saved UTF-8 text
a101f2218dc92de87299eaba40064a3e66f96120b89efd7b1490f6ef16a8d43e

Counts use a deterministic parser, not a spec-compliance grade. Unusual Markdown and indented code can fall outside its recognized patterns. See the llms.txt proposal and tokenizer implementation.

Related tools and resources

Contact Jon