Can ChatGPT, Claude or Gemini make an MP3 voiceover?

Yes, once Lazybird is added. Your agent creates the MP3 in the voice you choose and gives you the file.

Can I download ChatGPT's Read aloud audio?

No. Read aloud plays the reply and does not save audio. To get an MP3, add Lazybird to your agent.

You can create a free account. Speech uses the characters in your plan, and the free plan includes some each month.

Sign in for chat apps and CLIs. Coding agents can use an API key from lazybird.app/platform instead.

USE WITH AI AGENTS

Voiceovers for ChatGPT, Claude, Gemini and other AI agents

Add Lazybird to your AI agent once, then ask for an MP3 in the voice you pick. Or paste a script into the generator and hear it now, free.

Choose your agent

Which AI agent do you use?

Pick yours and follow the steps below.

Add to ChatGPT

Add to Claude

Other agents

Set up Lazybird for your agent

Real recordings of each setup, start to finish. It takes about a minute, and you sign in to your Lazybird account once.

Add Lazybird to ChatGPT

1
Install the plugin
Open Lazybird in ChatGPT and click Install plugin.
2
Sign in and allow
Sign in to Lazybird, or create a free account, then click Allow.
3
Ask for a voiceover
Type @Lazybird in any chat, then the voice and the script, for example: "Read this as an MP3 in a warm, friendly American voice: …". You get a player and a Download MP3 link.

Recorded in ChatGPT on a free account: install, sign in, then an MP3 from one message.

Add to ChatGPT

Add Lazybird to Claude

1
Open connectors
In Claude, open Settings, then Customize and Connectors. Click Add, then Add custom connector.
2
Paste the server
Enter the name Lazybird and the address https://api.lazybird.app/mcp (copy it from the box under the steps), then click Continue and Add. Claude detects the sign-in on its own.
3
Connect
Click Connect, sign in to Lazybird and click Allow.
4
Ask for a voiceover
Ask in plain words, for example: "Find me a calm British narrator voice on Lazybird", then "Use Arthur and read this as an MP3: …". If Claude asks before using Lazybird, click Allow. The answer links to your MP3.

Text

1https://api.lazybird.app/mcp

Recorded on claude.ai: adding the connector, then finding a voice and getting the MP3.

Add Lazybird to Gemini

1
Open connected apps
On gemini.google.com, open Settings at the bottom, then Connected Apps.
2
Add a custom app
In Custom apps, click Add a custom app, enter https://api.lazybird.app/mcp (copy it from the box under the steps), then click Next.
3
Sign in
Sign in to Lazybird when asked and click Allow.
4
Ask for a voiceover
Type @ in a chat, pick Lazybird, then ask, for example: "Read this script in a warm, friendly US English voice."

Text

1https://api.lazybird.app/mcp

Custom apps are available to personal Google accounts in the US, 18 and over, with Keep Activity on, in English.

Other AI agents and coding agents

1
Remote MCP
Any app or agent that supports remote MCP servers with OAuth sign-in: add the server address below and sign in to your Lazybird account.
2
Coding agents: use an API key
Create a key at lazybird.app/platform and give it to your agent. It sends the key in the X-API-Key header, as in the example below.

Terminal

1curl https://api.lazybird.app/v2/generate-speech \2  -H "X-API-Key: YOUR_API_KEY" -H "Content-Type: application/json" \3  -d '{"voiceId": "tgJ2Hes6DauwTg8kFrRPLN", "text": "Hello from my agent."}' -o hello.mp3

Get an API key

What your agent can do

Any app that supports remote MCP servers with OAuth sign-in can use Lazybird. Add https://api.lazybird.app/mcp and sign in to your Lazybird account. The assistant gets five tools:

Search voices: find a voice by words, language or gender, including your own cloned and designed voices.
Generate speech: turn up to 5,000 characters into an MP3 in the voice you pick, with speed, accent and an acting direction.
Design a voice: hear a new voice made from a description. It isn't kept until you save it.
Save a voice: keep a designed voice, or clone one from a clear 10 to 60 second recording of one speaker. Only clone voices you have the right to use.
Check your account: your plan, the characters used and left, and when they reset.

For Codex, Cursor and other agents that read SKILL.md files, the Lazybird skill shows the agent how to use these tools and the API:

Terminal

1npx skills add lazybirdai/skills

Customer reviews

Hear what people say about Lazybird

AndriaBauneeEnterprises

Verified buyer

Verified

★★★★★

Voice cloning is AMAZING

“The program itself takes a few tries to get used to. I was really interested in the voice cloning. One five-minute audio of me reading a book and BAM! minutes later it was loaded. I played with the expressiveness and speed and I CAN'T BELIEVE IT SOUNDS JUST LIKE ME!!!!! You can do multiple clones, so I am going to make a recording for different genres. This is over the top exciting. THANK YOU!”

Voice cloningRead review

DSMagic

Plus

Verified buyer

Verified

★★★★★

The best I've seen so far

“I've been using Googles studio and this is much better. I love it. The clone worked fantastic.”

Voice cloningRead review

Teoman38620

Verified buyer

Verified

★★★★★

Makes creating podcasts easy

“Lazybird uses your scripts (dialogues) to create voiceovers. The voiceovers can be one or more persons, though I only created two-person podcasts, a host and a guest...”

Podcast creationRead review

Nueirat

Verified buyer

Verified

★★★★★

Awesome: Effortless Script Handling

“I have tried several systems before, and I can confidently say this is the best for long scripts. I even used it for a long audio script that lasts around 75 minutes.The great thing is that you just need to bring the text and paste it in the first section; the system will take all the script, no matter how long, and just past it, and it will divide it into sections. There's no need to copy one paragraph or sentence at a time...”

Long scriptsRead review

Featured onAppSumoSee other reviews

Frequently asked questions

Start free

Make your first voiceover from a chat.

Create a free account, add Lazybird to your agent, and ask for a voiceover.

Create a free account Browse voices