Send audio from your phone with Telegram

Dimmy can watch your own Telegram account for audio. Record a voice note on your phone with any app, share it to your Telegram Saved Messages, and Dimmy on the PC picks it up, transcribes it, and writes a recap. It uses your own account, no bot and no cloud drive in between.

How it works

1. 1On your phone, record audio with any recorder or voice-note app.
2. 2Share or forward that audio into your Telegram Saved Messages (your own self-chat).
3. 3With Dimmy running on your PC, it notices the new audio, transcribes it, and generates a recap.

Connect Telegram (one-time setup)

Open Dimmy → Settings → Integrations → the Telegram card, and turn on the Telegram toggle. You log in with your own Telegram account, the same one you use on your phone. There is no bot to create and no visit to my.telegram.org.

1. 1Enter your Telegram phone number, with country code (for example +39 333 1234567). Click Send code.
2. 2Telegram sends a login code to your other logged-in devices (your phone, Telegram Desktop). Type it into the code field. Click Confirm.
3. 3If you have two-step verification on, Dimmy asks for your Telegram password. Enter it and click Confirm.
4. 4On success the card shows a green check and "Connected", or "Connected as <your name>" once it reads your account.

What Dimmy picks up

Dimmy watches one place only: your Telegram Saved Messages. Anything you send or forward to any other chat is ignored. In Saved Messages it acts on audio, and only audio.

• Voice notes recorded in Telegram itself.
• Audio files shared or forwarded to Saved Messages: .ogg, .oga, .opus, .m4a, .mp3, .wav, .aac, .flac, .webm, and similar audio formats.
• Up to Telegram's own ceiling of 2 GB per file.

When new audio arrives

By default a small window slides in at the bottom-right, titled "New audio from Telegram". It shows the file name and size and asks whether to transcribe and recap it, with two buttons: Dismiss, and Transcribe + recap. It does not steal focus and does not disappear on its own, so you can act on it whenever you get back to the PC.

Process every audio automatically

If you would rather skip the prompt, turn on "Transcribe and recap automatically" on the Telegram card (it appears once you are connected, and is off by default). With it on, every audio shared to Saved Messages is transcribed and recapped on its own, no click needed.

Where your transcript and recap land

The transcript is transcribed with the same engine as the "Transcribe a file" card and is always saved to your History. Then Dimmy generates a recap the same way it does for meetings, with a recap and action items, and the entry shows up alongside your meeting recaps.

• Success shows a "Telegram audio ready" toast once the file is transcribed and recapped.
• If the recap cannot be generated (for example your LLM is not set up), the transcript is still saved to History and you get a "Telegram audio transcribed" toast instead.
• If transcription itself fails, the item stays in your inbox to retry, so nothing is lost.

If your PC is off

Saved Messages live in Telegram's cloud, so audio you share while the PC is off simply waits. The next time you launch Dimmy it catches up and processes whatever arrived in the meantime. One honest limit: audio you shared before you first connected the account is not fetched retroactively. Only what arrives after you connect is picked up.

Privacy

• Read-only. Dimmy never sends messages, never replies into Telegram, never joins groups or adds contacts. It only reads audio from your Saved Messages.
• The audio is downloaded locally into Dimmy's data folder on this PC.
• Transcription and recap follow your normal STT and LLM settings, exactly like a regular dictation or meeting.
• Your Telegram session is stored locally on this device. Log out any time with Log out, and Dimmy forgets it.