Skip to content

Loquela Remote changelog

io.github.koaque/loquela-remote

3 changes recorded. MCP Harbor re-reads this listing from the official registry every six hours and asks the live server what it exposes, and keeps what changed between one look and the next.

  • server.json 2 fields changed
    description
    Private English transcription with speaker labels and word timings. US beta; paid minutes required. Private speech-to-text: English plus 24 European languages, speaker labels and word timings.
    version
    1.0.0 1.1.0

  • llms.txt +2

    https://api.getloquela.com/llms.txt

    + - PGP encryption (Pro, Business, Enterprise): audio encrypted to the account's own key (account page or GET /v1/pgp/key), transcripts encrypted to the customer's key (encrypt_to, or an encrypted-only API key) and signed by our key (https://api.getloquela.com/pgp-key.asc), so the network in between carries only ciphertext. MCP results are plain text.+ - [Partner program](https://api.getloquela.com/partners): Earn 20% of what referred customers pay in their first 12 months. Links without cookies; for people and businesses in the United States.

  • llms.txt 1 field changed
    file
    absent published

    https://api.getloquela.com/llms.txt

    + # Loquela Remote+ > Private, OpenAI-compatible speech-to-text API and MCP server for English audio. It runs on hardware we own in North Carolina; audio is processed in memory and never stored or used for training. For customers in the United States. In beta.+ - Speed: over 400× faster than real time, an hour of audio in under 9 seconds on our server, not counting upload (our own test, 27 October 2026).+ - Accuracy: 1.8% of words wrong on clear, non-noisy English speech (LibriSpeech test-clean), in our own test.+ - API: OpenAI's audio transcription format at https://api.getloquela.com/v1 (json, text, srt, vtt, verbose_json with word timings, diarized_json with speakers). Speaker labels cost nothing extra.+ - MCP server: https://api.getloquela.com/mcp, OAuth sign-in; tools transcribe a link, a short clip or a file sent to a one-time upload link.+ - Limits: up to 180 minutes of audio and 100 MB per request.+ - Prices (USD): Starter $9.99 a month for 5,000 minutes; Pro $95 a month for 50,000 minutes; Business $449 a month for 250,000 minutes; 2 months free when paid yearly; pay as you go $0.0035 a minute; Enterprise from $0.0008 a minute by agreement.+ - Contact: info@getloquela.com (Enterprise, security questionnaires, questions).+ ## Pages+ - [Pricing](https://api.getloquela.com/pricing): Plans, pay as you go, Enterprise and a dated comparison with other transcription APIs.+ - [Questions](https://api.getloquela.com/faq): Storage, training, where it runs, OpenAI compatibility, MCP clients, formats, speed, accuracy and billing.+ - [Meeting notes](https://api.getloquela.com/use-cases/meeting-notes): Transcribing meetings with speaker labels and word timings.+ - [Call recordings](https://api.getloquela.com/use-cases/call-recordings): Transcribing sales, support and interview calls without keeping the audio.+ - [Podcasts and video](https://api.getloquela.com/use-cases/podcasts-and-video): Transcripts and SRT or VTT captions for episodes and videos up to 3 hours.+ - [AI agents](https://api.getloquela.com/use-cases/ai-agents): Giving Claude, ChatGPT, Cursor or your own agent speech-to-text through MCP.+ - [How your audio is handled](https://api.getloquela.com/privacy-design): Each step a recording takes, what is never kept and what is.+ - [Security](https://api.getloquela.com/security): How audio, accounts and keys are protected, and how to report a problem.+ ## Docs+ - [API reference](https://docs.getloquela.com/remote): Endpoints, options, response formats, errors and limits.+ - [MCP setup](https://docs.getloquela.com/remote/mcp): Connecting Claude, ChatGPT, Cursor and other MCP clients.+ - [Apps](https://docs.getloquela.com/remote/apps): Using an API key in apps that accept an OpenAI-compatible transcription endpoint.