F
licenseNot graded
qualityD
maintenanceEnables AI agents to generate and play high-quality text-to-speech audio using the Kokoro model, with support for multiple voices, adjustable speaking speed, and audio caching.
-
realtime-tts-mcp
by tsushanth
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@realtime-tts-mcpConvert 'Welcome to our podcast' to speech using a female voice."
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
realtime-tts-mcp
An MCP server wrapping ReadAloud's realtime streaming text-to-speech API (Kokoro-82M) as a single synthesize_speech tool. Returns a playable WAV file. Typical warm latency is well under a second.
Tool
synthesize_speech(text, voice?, speed?)— converts text to spoken audio.voicedefaults toaf_heart(a Kokoro voice id);speeddefaults to1.0.
Related MCP server: Kokoro MCP Server
Install
npx realtime-tts-mcpOr from source:
git clone https://github.com/tsushanth/realtime-tts-mcp.git
cd realtime-tts-mcp
npm install
npm run buildAdd to your MCP client config:
{
"mcpServers": {
"realtime-tts": {
"command": "npx",
"args": ["-y", "realtime-tts-mcp"],
"env": {
"REALTIME_TTS_API_KEY": "rtts_your_key_here"
}
}
}
}Getting an API key
Self-serve signup at https://readaloudai.org/developers — sign in with email, generate a key. 10,000 characters free, no card required. Beyond that, pay-as-you-go at $0.01 per 1,000 characters (no plan to manage). Without a key, calls fail with a clear error pointing you to the signup page.
How it works
Each call does two steps: authorize (a fast key + free-tier/billing check against the ReadAloud API) then connects directly to the synthesis worker with a short-lived token — no relay hop in between, which is what keeps warm-call latency well under a second. A cold worker (idle for a couple minutes) can take several seconds longer on the first call after a gap.
Environment variables
Variable | Default | Purpose |
| (none, required) | Get one free at https://readaloudai.org/developers |
|
| API base URL |
|
| Per-synthesis timeout |
License
MIT
Available Tools
1 toolsynthesize_speechA
Convert text to spoken audio using a real-time, streaming Kokoro-82M TTS gateway. Returns a playable WAV file. The gateway currently runs on an unauthenticated CPU fallback (multi-second latency) — no API key required, but expect a few seconds per call.
ParametersJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| text | Yes | Text to synthesize | |
| speed | No | Playback speed multiplier (default 1.0) | |
| voice | No | Kokoro voice id (default "af_heart") |
TDQS
A4.2/5.0
Behavior4/5
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full behavioral disclosure burden. It proactively discloses the unauthenticated CPU fallback, latency expectations, and output format. It does not mention rate limits or text-length constraints, but the core runtime behavior and authentication requirements are transparent.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Conciseness5/5
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences with no filler. The primary action is front-loaded, and the latency/auth caveat is delivered concisely. Every sentence adds either functional or operational value.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Completeness4/5
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a three-parameter TTS tool with no output schema, the description covers the essential context: what it does, what it returns, and current performance constraints. It could mention voice-id availability or input size limits, but the schema provides the parameter structure and the description provides sufficient operational context.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Parameters3/5
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the baseline is 3. The description does not add any parameter-specific meaning beyond the schema; it focuses on overall tool behavior. Since the schema already documents text, speed, and voice, the lack of additional detail is acceptable.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Purpose5/5
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a specific verb and resource: 'Convert text to spoken audio using a real-time, streaming Kokoro-82M TTS gateway.' It also states the concrete output ('Returns a playable WAV file'), making the tool's function unmistakable even without sibling tools.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Usage Guidelines4/5
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear operational context: it is unauthenticated, requires no API key, and has multi-second latency. This implicitly tells an agent when use is acceptable (when latency is tolerable and no auth is available), though it does not explicitly discuss alternatives or exclusions since no siblings exist.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
1 tool update
v0.1.0- First observed
synthesize_speech
TDQS
A4.1/5.0
Scored across 1 tool
Disambiguation5/5
With only one tool, there is no possibility of overlap or confusion. synthesize_speech clearly indicates its single purpose of converting text to audio.
Naming Consistency5/5
The tool name follows a clear verb_noun pattern and directly matches the server's TTS purpose. There are no conflicting naming conventions to assess.
Tool Count3/5
A single tool is reasonable for a narrowly scoped TTS server, but the surface feels thin because related capabilities like voice selection or streaming control are absent.
Completeness3/5
The core text-to-speech operation is covered, but the toolset lacks visible voice, format, or streaming options that would be expected in a realtime TTS service. It works for basic synthesis but leaves notable gaps.
Maintenance
ActivityMaintained
ResponsivenessNo issues
Related MCP Connectors
Audio for your agent: transcribe, speak, translate, summarise, plus sound effects and music.
Speech, transcription, voice agents, Trace, Recap, dubbing and narration with browser OAuth.
AI voice generation: text-to-speech and voice cloning from any MCP client.
Turn any LLM multimodal; generate images, voices, videos, 3D models, music, and more.
Related MCP Servers
- AlicenseNot gradedqualityDmaintenanceProvides text-to-speech generation using the Kokoro-82M model, enabling AI assistants to generate voiceovers and audio content directly within Claude Desktop and Cursor.14Apache 2.0
leanvox-mcpofficial
AlicenseNot gradedqualityDmaintenanceEnables text-to-speech generation, voice cloning, dialogue creation, and other TTS operations through natural language in MCP-compatible AI assistants.7 npmMIT- AlicenseAqualityBmaintenanceEnables AI agents to generate speech using Gemini TTS models, with tools for text-to-speech, task polling, and pricing checks.439 npmApache 2.0