About
About Scriptivox
Scriptivox is an AI transcription company. We turn recorded audio and video into accurate, structured text — for people who need a readable transcript of a meeting or an interview, and for developers who need transcription inside their own product.
The problem we set out to solve is a narrow one, and we would rather do it properly than do many things adequately. Speech-to-text is easy to demo and hard to ship: real recordings have several people talking over each other, background noise, accents, code-switching between languages, and file formats that were designed decades apart. A transcript that is 90% right is not usable, because finding the wrong 10% costs more than typing the whole thing. So the work is in the parts nobody demos — speaker diarization that does not merge two people into one, word-level timestamps that survive editing, honest confidence scores, and failure modes that tell you what went wrong instead of silently returning something plausible.
Everything we build sits on that one engine. The web app, the API and the AI-agent integrations are three ways into the same transcription pipeline, so a transcript you get from an agent is the same transcript you would get from the dashboard.
What we build
Transcription web app
Upload audio or video at scriptivox.com and get an editable, speaker-labelled transcript with word-level timestamps. Export as plain text, SRT, WebVTT, DOCX or PDF, or edit captions directly in the browser.
Developer API
The same transcription engine over a REST API at api.scriptivox.com/v1, with presigned uploads, HMAC-signed completion webhooks, cursor-paginated job listing and caption export. Billed per hour of audio rather than per seat.
Free browser tools
Audio and video converters, joiners, trimmers and a subtitle editor that run entirely in your browser using WebAssembly. Files never leave your device for these tools, and no account is required.
AI agent integrations
A Model Context Protocol server published on npm as @scriptivox/mcp-server, so assistants like Claude and ChatGPT can transcribe files as a native tool call, plus an OpenAPI specification describing the whole API surface.
How we handle your files
Source audio you upload for transcription is deleted 24 hours after the job ends. The transcript itself stays available to you until you delete it, and deleting a transcription removes the stored transcript immediately. The browser-based conversion and editing tools do not upload anything at all — they run on your own machine through WebAssembly.
The full detail, including what we log and who processes it, is in the privacy policy. Our commitments and yours are in the terms of use.
At a glance
- Founded
- 2024
- Languages supported
- 119
- Maximum file size
- 5 GB
- Maximum duration
- 10 hours
- API pricing
- $0.20 per hour of audio
- Source audio retention
- 24 hours after a job finishes
Getting in touch
Support questions go to support@scriptivox.com, and anything commercial to sales@scriptivox.com. Both are answered in English. There is also a contact form.
Live uptime and incident history for the API and the web app are published at status.scriptivox.com. If you are building against the API, start with the developer documentation.