diff --git a/README.md b/README.md
index e910ce7..dd686b6 100644
--- a/README.md
+++ b/README.md
@@ -311,7 +311,7 @@ Set a custom full transcription endpoint when your provider accepts the OpenAI-c
 |----------|---------|---------|
 | `TELEPI_TRANSCRIPTION_API_KEY` | `OPENAI_API_KEY` | Credential sent to the provider |
 | `TELEPI_TRANSCRIPTION_URL` | OpenAI's `/v1/audio/transcriptions` endpoint | Full HTTP(S) endpoint |
-| `TELEPI_TRANSCRIPTION_MODEL` | `whisper-1` | Value for the multipart `model` field |
+| `TELEPI_TRANSCRIPTION_MODEL` | `gpt-transcribe` | Value for the multipart `model` field |
 | `TELEPI_TRANSCRIPTION_AUTH_HEADER` | unset | Custom credential-header name; unset uses `Authorization: Bearer` |
 | `TELEPI_TRANSCRIPTION_PROMPT` | unset | Optional recognition context for project-specific names and terms |
 
diff --git a/src/voice.ts b/src/voice.ts
index 27e6413..8694cb9 100644
--- a/src/voice.ts
+++ b/src/voice.ts
@@ -66,7 +66,7 @@ const TRANSCRIPTION_API_KEY_ENV = "TELEPI_TRANSCRIPTION_API_KEY";
 const TRANSCRIPTION_AUTH_HEADER_ENV = "TELEPI_TRANSCRIPTION_AUTH_HEADER";
 const TRANSCRIPTION_PROMPT_ENV = "TELEPI_TRANSCRIPTION_PROMPT";
 const DEFAULT_TRANSCRIPTION_URL = "https://api.openai.com/v1/audio/transcriptions";
-const DEFAULT_TRANSCRIPTION_MODEL = "whisper-1";
+const DEFAULT_TRANSCRIPTION_MODEL = "gpt-transcribe";
 const FFMPEG_INSTALL_MESSAGE = `ffmpeg not found. Install it with: ${getPlatformInstallHint("ffmpeg")}`;
 const NO_BACKEND_ERROR = `Voice messages require a transcription backend.