Works insideChatGPT
Transcribe audio and video inside ChatGPT
Add Whisper Web to ChatGPT as a custom app with OAuth. Then ask in plain language for transcripts, quotes, summaries or subtitles of your recordings and media links.
- ChatGPT web
- OAuth sign-in, no API key
- 99 languages
- Files up to 5 GB · 10 h
Illustrative example · tool names match the live server
01 / Why ChatGPT + Whisper Web
Bring your recordings into ChatGPT.
ChatGPT can summarize, rewrite and translate. Whisper Web gives it accurate, timestamped text to work from.
01
A custom app with OAuth
Create the app once from ChatGPT's settings. ChatGPT discovers Whisper Web's sign-in itself: no client ID, secret or API key.
02
Timestamps it can cite
Results include segment timestamps, so ChatGPT can quote the exact minute and build SRT or VTT subtitles from the same transcript.
03
An uploader in the conversation
ChatGPT can show Whisper Web's uploader inside the chat. Files go from your browser to Whisper Web storage, not through the chat.
02 / Setup
Connect ChatGPT in four steps.
ChatGPT adds MCP servers as custom apps in Developer mode. It takes about two minutes.
- Add API creditsOpen the Developer Console with your Whisper Web login and buy a credit pack (from $10). MCP bills $0.01 per audio minute, by the second.
- Turn on Developer modeIn ChatGPT, open Settings → Apps → Advanced settings and turn on Developer mode.
- Create the appChoose Create app, name it Whisper Web, paste the server URL and pick OAuth. Sign in to Whisper Web and approve access.
- Use it in a chatPick Whisper Web from the tools menu in a chat, then ask. Every transcription is also listed in your Developer account.
ChatGPT
Add Whisper Web as a custom app with OAuth.
https://whisperweb.dev/mcp- Transport
- Streamable HTTP
- Auth
- OAuth 2.1 · PKCE
- Billing
- $0.01 / audio min
- Scopes
- read · write · usage
Developer mode is available on ChatGPT web for eligible plans. In Business and Enterprise workspaces, an admin may need to allow custom apps first.
Using Claude instead? See the Claude MCP guide. Claude Code, Codex and Cursor setup is on the Whisper MCP page.
03 / Just ask
One sentence per job.
No commands to learn. Describe what you want and ChatGPT picks the Whisper Web tool.
Write
Turn recordings into finished text.
- “Transcribe this webinar and write a blog post from it.”
- “Summarize this interview in five bullet points with timestamps.”
- “Draft a LinkedIn post from the best two minutes of this talk.”
Find
Ask the recording questions.
- “What did the guest say about hiring?”
- “List every action item mentioned in this meeting.”
- “Where does the speaker first mention the budget?”
Ship
Get files out the door.
- “Give me VTT subtitles for this lecture.”
- “Transcribe this French interview and translate it to English.”
- “Show my last five transcriptions.”
04 / Pricing
$0.01 per audio minute.
Prepaid Developer API credits, billed by the second. Connecting, estimating, checking status and reading results are free.
$10
1,000 audio minutes
$25
2,500 audio minutes
$100
10,000 audio minutes
One-time packs, no subscription or automatic top-ups. Website Unlimited is a separate plan and does not include API credits. Buy credits
05 / Privacy & control
Your audio goes to Whisper Web, not to OpenAI.
01
Sign in, no keys
OAuth 2.1 with PKCE. Access tokens are bound to whisperweb.dev/mcp and expire after an hour.
02
Deletion is opt-in
Reading, transcribing and usage are granted by default. Deleting results needs its own permission.
03
Disconnect anytime
See every connected agent in Developer Console → AI agents (MCP). After you disconnect one, its next call is refused.
04
Uploads go to Whisper Web
The in-chat uploader sends files from your browser to Whisper Web storage, not through the assistant. Transcription uses cloud processing.
Only transcribe media you own or are authorized to use. Read the API data guide.
06 / FAQ
ChatGPT MCP FAQ
What is the Whisper Web ChatGPT app?
It is the Whisper Web MCP server at https://whisperweb.dev/mcp, added to ChatGPT as a custom app. ChatGPT can then transcribe audio and video, check costs and read transcripts with your Whisper Web Developer account.
How do I add it to ChatGPT?
Open Settings → Apps → Advanced settings and turn on Developer mode. Choose Create app, paste the server URL, pick OAuth and sign in to Whisper Web. Then select Whisper Web from the tools menu in a chat.
Do I need a paid plan or an API key?
You need a Whisper Web Developer account with prepaid API credits: $0.01 per audio minute, billed by the second. There is no API key to copy, because you sign in with OAuth. The website's Free and Unlimited plans do not include API credits.
How do I transcribe a file from my computer?
Ask ChatGPT to upload a file and pick it in Whisper Web's uploader in the chat; it goes from your browser to Whisper Web storage. Each file can be up to 5 GB and 10 hours.
Can it transcribe YouTube links?
Yes. Paste a single YouTube, Facebook or X (Twitter) video link, or a direct audio or video URL. Only transcribe media you own or are authorized to use, and respect the platform's terms and the creator's rights.
Does my audio go to OpenAI?
Files picked in the uploader go from your browser to Whisper Web, not through ChatGPT. Transcription runs in Whisper Web's cloud. ChatGPT receives the transcript text it asks for, which then becomes part of your conversation under OpenAI's terms.
How do I disconnect ChatGPT?
Open Developer Console → AI agents (MCP) and choose Disconnect, and its next call is refused. You can also remove Whisper Web in ChatGPT under Settings → Apps.
Transcribe in ChatGPT today.
Add credits, connect once and send your first recording without leaving the chat.