Install Jentic One Beta
Jentic One is a self-hosted execution layer for AI agents. It lets your agent call the Asynchronous Speech-To-Text API Documentation, or any other public or private API you need. You set the rules, the agent never sees your credentials, and every call is logged.
Two steps, two machines. Install the instance in a safe environment, then register your agent from wherever it runs.
Step 1: Jentic One Host machine
# On the machine that will host your Jentic One instance:
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frev.ai%2Frev" | shStep 2: Agent machine
# On the machine where your agent runs (keep this separate from the instance):
curl -fsSL "https://jentic.com/install.sh?src=apis&api=%2Fapis%2Frev.ai%2Frev" | sh
jentic register # connects your agent to your Jentic One instanceJentic One is in public beta. The setup above keeps your agent separate from the instance, which is what you want before using real credentials: an agent running as the same OS user as Jentic One can read its stored keys directly. Just evaluating? A single local install is fine to start. See the secure deployment guide for the tiers.
What an agent can do with Asynchronous Speech-To-Text API Documentation API.
Start a transcription job from a public media_url or a multipart media upload of under 2GB with POST /jobs
Get notified at a callback_url when a job finishes instead of polling
Fetch a finished transcript as speaker monologues in JSON or as plain text from /jobs/{id}/transcript
Download captions in SubRip (SRT) or WebVTT format from /jobs/{id}/captions
GET STARTED
Tune a job with language, skip_diarization, skip_punctuation, filter_profanity, remove_disfluencies, and custom_vocabularies
Check remaining balance_seconds on the account with GET /account
Permanently delete a completed job and its media and transcript with DELETE /jobs/{id}, or set delete_after_seconds up front
Patterns agents use Asynchronous Speech-To-Text API Documentation API for, with concrete tasks.
★ Agent-run meeting transcription through Jentic
An assistant agent connected through Jentic searches by intent for transcribing a recording, gets SubmitTranscriptionJob with its schema, and submits the meeting's media_url with a callback_url. When the job status becomes transcribed, the agent fetches the JSON transcript with its speaker monologues and writes up a summary with who said what.
POST /jobs with media_url 'https://example.com/meeting.mp3', language 'en', and callback_url 'https://example.com/hooks/rev', then GET /jobs/{id}/transcript once the job status is transcribed
Captions for recorded video
A video publishing pipeline submits each upload to Rev.ai, then requests captions from /jobs/{id}/captions with an Accept header of text/vtt for web players or application/x-subrip for SRT. Captions default to SRT, and speaker_channel lets you request captions for one channel of a multichannel recording.
POST /jobs with media_url 'https://example.com/webinar.mp4', then GET /jobs/{id}/captions with Accept 'text/vtt' once the job is transcribed
Domain vocabulary and privacy controls
For medical, legal, or product recordings full of unusual terms, a job can pass custom_vocabularies or a custom_vocabulary_id to improve recognition of those words. Setting delete_after_seconds, or calling DELETE /jobs/{id} after download, removes the media and transcript from Rev.ai so sensitive recordings are not retained.
POST /jobs with media_url 'https://example.com/consult.wav', custom_vocabularies listing the product names, and delete_after_seconds 86400
Job tracking and failure handling
GET /jobs lists jobs from the last 30 days in reverse chronological order, paged with limit and starting_after. For any job whose status is failed, the failure field gives a reason code such as invalid_media, duration_exceeded, or insufficient_balance, with failure_detail explaining it, so an agent can resubmit or alert you.
GET /jobs with limit 50, then for each job whose status is failed report its failure and failure_detail values
7 endpoints — the rev.
METHOD
PATH
DESCRIPTION
/jobs
Submit a transcription job
/jobs/{id}
Get job status and details
/jobs/{id}/transcript
Get the transcript
/jobs/{id}/captions
Get SRT or VTT captions
/jobs
List recent jobs
/jobs/{id}
Delete a job and its data
/account
Get account balance
/jobs
Submit a transcription job
/jobs/{id}
Get job status and details
/jobs/{id}/transcript
Get the transcript
/jobs/{id}/captions
Get SRT or VTT captions
/jobs
List recent jobs
/jobs/{id}
Delete a job and its data
/account
Get account balance
What agents get from Jentic-routed access to this vendor.
Setup
Wiring Rev.ai by hand means handling its bearer access token, choosing between JSON and multipart job submission, setting the right Accept header for transcript and caption formats, and running callbacks or polling. Through Jentic you install once, import the Rev.ai API from the API Directory, store the token once, and your agent calls it.
Permission scoping
Rev.ai puts the job id in the URL path (/jobs/{id}/...), so a rule can limit your agent to reading one job's transcript and captions. You choose the operations it may call, so destructive ones like DeleteJobById are not included unless you add them.
Credential isolation
Your Rev.ai access token is stored once, encrypted, by your own Jentic One instance and injected at execution time. It never enters the agent's prompt, logs, or context.
Intent-based discovery
Agents search Jentic by intent such as 'transcribe an audio file' or 'get captions for a video', and Jentic returns the matching Rev.ai operation with its input schema so the agent calls the right endpoint without browsing the reference docs.
Alternatives and complements available in the Jentic catalogue.
Specific to using Asynchronous Speech-To-Text API Documentation API through Jentic.
What authentication does the Rev.ai Speech-To-Text API use?
Per its OpenAPI spec, requests carry an access token as a bearer token in the Authorization header. You generate the token once on your Rev.ai settings page, and the spec notes it does not expire. Through Jentic the token is stored encrypted by your own Jentic One instance and injected at call time.
Which languages can the Rev.ai Speech-To-Text API transcribe?
The language option on POST /jobs accepts 31 language codes in the spec, such as en, es, fr, de, ja, cmn, and hi. The spec notes that media longer than 17 hours is not supported for English and longer than 12 hours for other languages, and non-English jobs can take up to 6 hours.
What are the rate limits for the Rev.ai Speech-To-Text API?
The OpenAPI spec does not specify rate limits. Check https://docs.rev.ai for current limits. The spec does cap direct uploads at under 2GB, so larger files should be submitted by media_url.
Is there a Rev.ai MCP server?
You don't need an MCP server to give your agent the Rev.ai Speech-To-Text API. Jentic connects it directly from the API Directory: import it, store your access token once, and your agent calls it. Operations are found by intent when needed, so no extra tool definitions sit in your agent's context.
How do I transcribe an audio file through Jentic?
Install Jentic One, add the Rev.ai Speech-To-Text API from the directory, and store your access token. Your agent searches for 'transcribe an audio file', gets POST /jobs with its schema, submits the media_url, then calls GET /jobs/{id}/transcript once the status is transcribed. To run it on your own infrastructure, install Jentic One from its GitHub repo.
Can I limit what my agent is allowed to do with the Rev.ai Speech-To-Text API?
Yes. Write a rule that allows only POST /jobs, GET /jobs/{id}, and GET /jobs/{id}/transcript, so the agent can submit recordings, check status, and read transcripts but cannot delete jobs or list your other jobs, and every call it makes is logged.
Know of an official OpenAPI document? Contribute it →
For Agents
Submit recorded audio or video for asynchronous transcription, poll or receive a callback when it is done, and fetch the transcript as JSON or text, or captions as SRT or VTT.
Use for: Transcribe this podcast episode from its MP3 URL, I need SRT captions for a recorded webinar, Check whether my transcription job has finished, List all transcription jobs from the last week
Not supported: Does not handle live streaming transcription, text-to-speech, translation, or human transcription orders. Use for asynchronous transcription of recorded audio and video only.
The Rev.ai Asynchronous Speech-To-Text API transcribes recorded audio and video files into text. You submit a job with a public media_url or a multipart file upload, Rev.ai processes it in the background and can call your callback_url when it finishes, and you then fetch the transcript as JSON or plain text, or captions as SRT or VTT. Job options cover language, speaker diarization, punctuation, profanity filtering, disfluency removal, custom vocabularies, and automatic deletion after a set number of seconds.
This API is usable in Jentic One now. Its AI-readiness score against Jentic's framework shows where it stands today and where improvements would make it even easier for agents to use.
Base layer of spec validity and structural soundness.
Aggregated quality score from linter diagnostics, weighted by severity.
Percentage of `$ref` references that resolve successfully.
Checks whether the API description parses successfully and conforms to its declared specification (e.g., OpenAPI).
Structural correctness score based on schema issues using logarithmic dampening.
Clarity, completeness, and ingestion readiness for developers and tooling.
How richly the API is illustrated with examples.
Percentage of examples that conform to their schemas.
Percentage of operations with complete response definitions (success, client error, server error).
Health of API ingestion, bundling, and resolution within Jentic pipelines.
Semantic breadth, depth, and agent comprehension for AI systems.
Coverage of descriptions across API elements.
Coverage of RFC 9457 Problem Details for error responses.
Coverage, uniqueness, and casing consistency of operationIds for AI inference.
Coverage of summaries across operations/tags/info.
Functional utility, complexity comfort, and AI orchestration readiness.
Agent comfort level based on API operational and structural complexity.
Trust, risk posture, and security compliance.
Average quality of security schemes based on authentication method strength (weakest link for OAuth2).
Findability, semantic richness, and reasoning readiness.
Clarity and depth of descriptions across API elements.
Score it yourself
Every API in the directory is allowlisted, so you can re-score it with no key required.
npx @jentic/api-scorecard-cli score <openapi-url>