Vapi (voice agent API) by Vapi
API to build and run phone and web voice agents with your choice of speech, model and voice providers.
Voice (speech-to-text, text-to-speech) Official MCP server
+ Build a stack with Vapi (voice agent API) Website Official pricing GitHub
How Vapi (voice agent API) compares
Vapi (voice agent API) is one of 12 APIs in the Voice (speech-to-text, text-to-speech) category on AgentGid. Its cheapest paid plan ($29 a month) is the 4th cheapest of the 6 here with a monthly plan. It is one of 11 that can be started for free. It ships an official MCP server, like 5 others here. Its official SDK is the 8th most downloaded in the category (140K a week). See the full table.
Pricing
| Free tier | $5 free credits to start |
|---|---|
| Cheapest paid plan | $29 a month |
| Usage price | $0.05/min Vapi hosting; models, speech and telephony billed at cost on top |
| Good to know | Vapi's own estimate for 1,000 minutes on its pricing page (as reported): hosting $50, Deepgram transcriber $10, OpenAI model $8-$45, ElevenLabs voice $15-$24, plus transport. Core plan $29/month adds concurrency and data retention. |
From the official pricing page, checked Oct 22, 2026. Prices change: confirm before you buy.
Use it from an agent
| MCP server | Official: @vapi-ai/mcp-server |
|---|---|
| npm | npm install @vapi-ai/server-sdk |
| PyPI | pip install vapi_server_sdk |
| GitHub | VapiAI/server-sdk-typescript · 122 stars · last push Oct 16, 2026 |
Alternatives to Vapi (voice agent API)
| API | Free | Price | SDK downloads / wk | GitHub stars | MCP |
|---|---|---|---|---|---|
| OpenAI Realtime API Speech-to-speech models for real-time voice agents, priced per 1M audio and text tokens. |
— | gpt-realtime-2.1: audio $32 in / $64 out per 1M tokens;... | 107M | 11.2K | |
| ElevenLabs Text-to-speech, speech-to-text (Scribe), voice cloning and conversational voice-agent APIs. |
Free plan: 10,000 TTS characters/month (v4) | Free plan · paid from $6/mo | 2.8M | 3,150 | MCP |
| LiveKit Deploy and host AI voice and video agents with real-time communication. |
Build plan: 1,000 agent session minutes, 1... | Free plan · paid from $50/mo | 2.8M | 469 | |
| Deepgram Real-time and batch speech-to-text, text-to-speech and voice-agent APIs. |
$200 credit, no credit card | Free plan then Nova-3 streaming $0.0048/min (promo; regular $0.0077);... |
1.4M | 469 | MCP |
| Daily Video SDK and WebRTC infrastructure for building calls and voice AI agents. |
10,000 free minutes/month | Free plan | 510K | — | |
| Retell AI Platform and API for phone voice agents and chat agents, billed per minute or message. |
$10 in free credits | Free plan then $0.07-$0.31/min for AI voice agents; $0.002+/msg for chat agents |
315K | 62 | MCP |