1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
//! Speech-to-text transcription backends.
//!
//! Defines the `SpeechTranscriber` trait — the single abstraction that
//! decouples the daemon from any particular inference runtime.
//!
//! Backend:
//! - `http` — delegates to an external server (llama-server, vLLM, etc.)
use Result;
/// Result of a transcription.
/// Contract that every STT backend must fulfil.
///
/// Implementations are expected to be long-lived — created once at startup
/// and called repeatedly for the lifetime of the daemon.