Audio Recordings

In-App Audio Recordings, Player, & Transcriptions

In the current version, audio recordings and files are not able to be deleted from chat once sent.

Audio Recordings with Transcription

Voice Messages

Vector Audio Recording & Audio Player

Voice messages and audio recordings are supported. Currently, .mp3 .wav .flac .aac .m4a .ogg .opus audio filetypes are supported and playable within Vector Messenger. A small audio player with basic functionality comes built-in. It displays the time duration of the audio file or recording with the ability to pause and play. Audio files are automatically downloaded and installed on the user's local device and can be found under the vector folder under Downloads.

Vector Messenger Audio Transcription

Audio Transcription (Spanish)

---

How to Record

Click Microphone to Record Audio

Click Microphone Again to Send

Vector Messenger audio and voice recordings function like most messenger apps. Instead of holding down the Microphone icon for the duration of your desired recording, tap it once to start and tap it again to send.

---

Audio Transcriptions (AI)

Vector Messenger now allows for users to transcribe or translate audio recordings into text. This feature is still experimental and not all transcriptions are 100% accurate, but it does have multi-language support. Currently, audio transcription formats (voice-to-text) that are supported include: .wav .mp3 .flac. Vector's design is primarily focused on privacy and for users to have full control of what they want and the ability to opt-in and opt-out. Therefore, under Settings you can find Voice Settings and turn on/off Auto-Translate, Auto-Transcribe, and even pick the custom AI model you want to download locally and run for your Audio Transcriptions. By running these LLMs (AI) locally, it is private and information is not shared externally without your consent. You can handpick the Whisper Model that best fits your need and can even delete or remove them to save virtual memory.

Vector Messenger Voice Settings (Settings)

Whisper Model Selection (AI)

Whisper Model (AI)

Whisper Model, aka Vector Voice AI Model, determines the quality of your transcriptions. A larger model will provide more accurate transcriptions and translations, but require more disk space, memory, and GPU power to run on your local device. You are trading computing resources for higher quality output. By default, the Base Quality model is selected, but a user may choose to remove (uninstall) or run a quality model better suited for their needs.

Auto-Translate

Vector Voice Translations—Vector Voice AI can automatically detect non-English languages and translate them in to English text for you. You can decide whether Vector Voice transcribes to your contact's native language or translates to English instead.

Auto-Transcribe

Vector Voice Transcriptions—Vector Voice AI can automatically transcribe incoming Voice Messages for immediate reading, without needing to listen. You can decide whether Vector Voice transcribes automatically or if you prefer to transcribe each message individually.