cd /news/ai-agents/gradium-mcp-beta · home › topics › ai-agents › article
[ARTICLE · art-172750] src=gradium.ai ↗ pub= topic=ai-agents verified=true sentiment=↑ positive

Gradium MCP beta

Gradium released a beta Model Context Protocol server at https://mcp.gradium.ai/mcp that lets coding agents in Claude Code and Codex browse Gradium's voice catalog and generate text-to-speech without leaving the session. The server exposes four tools: list_voices and get_voice are free, get_credits shows balance and plan, and generate_speech costs 1 credit per character at the same rate as Gradium's API, with sign-in through an existing Gradium account and no API key. Gradium also published four Claude Code and Codex skills in the gradium-ai/skills GitHub repository and said speech-to-text and voice design are the next candidates for the beta.

key takeaway

Gradium's MCP server at https://mcp.gradium.ai/mcp exposes list_voices, get_voice, generate_speech and get_credits to Claude Code and Codex, with voice browsing free and speech billed at 1 credit per character, matching the API rate.

read5 min views1 publishedNov 19, 2026
Gradium MCP beta
Image: Gradium (auto-discovered)

Text-to-Speech and voice selection from your coding agent

The Gradium MCP is available today in beta at https://mcp.gradium.ai/mcp. Ask your coding agent to turn a short PDF into an audiobook or add a voiceover to your video, without leaving your coding session.

We built the Gradium MCP to make exploring Gradium voices easier. You can sign in to your Gradium account from your agent, with no detour to Studio. The agent can then browse the catalog and recommend voices for you.

Set it up with a coding agent #

Paste this prompt into Claude Code or Codex. You need a Gradium account, and no API key.

Connect the Gradium MCP at https://mcp.gradium.ai/mcp, then find a friendly voice and read this aloud:
"I’ve got a parcel for Mr. Sherlock Holmes at 221B Baker Street. Could you look it up? The tracking code is RF24, and the surname is spelled H-O-L-M-E-S."

Your agent adds the server and asks you to sign in. If the Gradium tools don't show up after you sign in, start a new session and paste the prompt again.

What the Gradium MCP does #

Tool What it does Cost
list_voices Browses the Gradium catalog and your organization's custom voices, filtered by language, accent, gender or use case Free
get_voice Shows one voice's details, such as its accent, gender, age and description Free
generate_speech Turns your text into speech with the voice you choose 1 credit per character
get_credits Shows your credit balance and plan Free

For example, get_voice describes Everly as a middle-aged, female, American English (en-us) voice with a "rounded, soft-edged tone with gentle lift at the end of phrases."

list_voices lists the best-ranked voices first, starting with the flagship voices our team picked by hand. Your agent browses voices for free, and only uses credits when it generates audio, at the same rate as the API.

Add the Gradium MCP manually #

If you'd rather run the commands yourself, use these.

Claude Code

claude mcp add --transport http --scope user gradium https://mcp.gradium.ai/mcp

Then run /mcp in a session, select gradium and choose Authenticate. If you install the Gradium plugin, skip this command. The plugin already includes the server, and Claude Code connects each URL only once.

Codex

codex mcp add gradium --url https://mcp.gradium.ai/mcp
codex mcp login gradium

The Gradium skills for Claude Code and Codex #

The Gradium plugin adds four skills that help your agent use the MCP:

  • Setup connects to the MCP and explains sign-in errors.
  • Credits checks your balance and which organization to use.
  • Voice selection shortlists voices that match your brief.
  • Speech generation turns your script into audio files in your project.

You'll find the skills, with install steps for Claude Code and Codex, in the gradium-ai/skills repository.

What's coming to the Gradium MCP #

The beta covers Text-to-Speech and choosing voices. Speech-to-Text and voice design are the next candidates, and your feedback decides which ships first.

We're also submitting the Gradium plugin to the Claude Code and Codex plugin directories, so installing it becomes a single step.

Tell us how you use it #

Tell us on Discord:

  • what you're using the Gradium MCP for, or plan to use it for
  • which agent you use it with, and how the setup went
  • what worked well, and what got in your way
  • which tool or feature you want next

Get started #

The Gradium MCP works with your existing Gradium account and credits, with no API key to manage.

Jul 30, 2026ResearchDiscover

Gradium TTS public beta: open for your hardest cases

A new Gradium TTS model is available today in public beta. It keeps improving on the current production model, handling the complex cases natively, phone numbers, email addresses, IBAN numbers, time expressions and more, with no pre-processing or text normalization on your side. Anyone can test it through the API, and everyone who sends us feedback during the beta receives 1M Gradium credits.

Feb 11, 2026EngineeringDiscover

Optimizing Quality vs. Latency in Real-Time Text-to-Speech AI Models

Explore strategies for balancing quality and latency in real-time TTS AI models. Learn how Gradium achieves low-latency, high-quality speech synthesis for voice applications.

Frequently Asked Questions #

Yes. The beta covers Text-to-Speech and choosing voices, and we're adding more tools soon. Tell us on Discord what you'd use it for.

No. You sign in with your Gradium Studio account over OAuth when your client connects. The server never asks for an API key.

We've tested the Gradium MCP in Claude Code and Codex. In both, the Gradium plugin adds four skills on top of the MCP server.

Browsing voices, reading voice details and checking credits are free. generate_speech uses 1 credit per character of text, the same rate as the Text-to-Speech API. Auditioning a voice means generating audio with it, so auditions use credits too.

The catalog includes voices in English, French, German, Spanish and Portuguese, each led by a flagship voice. Your organization's custom voices work too.

generate_speech accepts up to 9,000 characters per call and returns at most about 2 minutes and 55 seconds of audio, roughly 2,000 characters. In Claude Code, the speech generation skill splits longer scripts into parts and can join them into one file.

They're the next candidates for the MCP. Today you can use Speech-to-Text, voice design and voice cloning through the API and in Studio.

The docs MCP searches the Gradium documentation. The Gradium MCP generates speech and browses voices in your account.

── more in #ai-agents 4 stories · sorted by recency
── more on @gradium 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
→ Live at https://your-agent.zahid.host ✓
Get free account → Pricing
from €0/mo · no card required
LIVE [news/gradium-mcp-beta] indexed:0 read:5min 2026-11-19 · —