fix(voice): scope OpenAI voice credentials

Resolve OpenAI voice credentials and endpoints from voice-specific config before broader OpenAI fallbacks, and use the same REST STT provider for dictation and voice mode.

docs/data-model.md

@@ -163,7 +163,7 @@ Single file, validated with `PersistedConfigSchema`.
     root?: string            // optional root for new worktrees; defaults to $PASEO_HOME/worktrees
   },
   providers: {
-    openai: { apiKey: string },
+    openai: { voice: { apiKey: string, baseUrl: string } },
     local: { modelsDir: string }
   },
   agents: {
@@ -193,6 +193,37 @@ All fields are optional with sensible defaults.
 
 Local speech model ids are intentionally narrow: STT uses `parakeet-tdt-0.6b-v2-int8`, TTS uses `kokoro-en-v0_19`, and turn detection uses the bundled Silero VAD model.
 
+Set these to select OpenAI instead of local speech:
+
+| Env var                        | Applies to                      |
+| ------------------------------ | ------------------------------- |
+| `PASEO_VOICE_STT_PROVIDER`     | Voice mode STT provider         |
+| `PASEO_DICTATION_STT_PROVIDER` | Composer dictation STT provider |
+| `PASEO_VOICE_TTS_PROVIDER`     | Voice mode TTS provider         |
+
+OpenAI voice can be configured under `providers.openai`:
+
+```json
+{
+  "providers": {
+    "openai": {
+      "voice": {
+        "apiKey": "sk-...",
+        "baseUrl": "https://api.openai.com/v1"
+      }
+    }
+  }
+}
+```
+
+`providers.openai.voice.apiKey` and `providers.openai.voice.baseUrl` apply only to Paseo OpenAI voice features.
+
+Paseo uses these paths under the configured OpenAI base URL:
+
+- dictation STT: `/v1/audio/transcriptions`
+- voice mode STT: `/v1/audio/transcriptions`
+- voice mode TTS: `/v1/audio/speech`
+
 ---
 
 ## 3. Schedule