From e111cbb36d945acc231961313051f1b2198d95bf Mon Sep 17 00:00:00 2001
From: Abhishek Kumar <abhishek@a6k.me>
Date: Fri, 20 Feb 2026 20:41:11 +0530
Subject: [PATCH] feat: add cartesia tts

---
 api/services/configuration/registry.py  | 20 ++++++++++++++++++++
 api/services/pipecat/service_factory.py |  8 ++++++++
 2 files changed, 28 insertions(+)

diff --git a/api/services/configuration/registry.py b/api/services/configuration/registry.py
index ab28a0d..7955a4b 100644
--- a/api/services/configuration/registry.py
+++ b/api/services/configuration/registry.py
@@ -277,6 +277,19 @@ class DograhTTSService(BaseTTSConfiguration):
     api_key: str
 
 
+CARTESIA_TTS_MODELS = ["sonic-3"]
+
+
+@register_tts
+class CartesiaTTSConfiguration(BaseTTSConfiguration):
+    provider: Literal[ServiceProviders.CARTESIA] = ServiceProviders.CARTESIA
+    model: str = Field(
+        default="sonic-3", json_schema_extra={"examples": CARTESIA_TTS_MODELS}
+    )
+    voice: str = Field(default="a167e0f3-df7e-4d52-a9c3-f949145571bd")
+    api_key: str
+
+
 SARVAM_TTS_MODELS = ["bulbul:v2", "bulbul:v3"]
 SARVAM_V2_VOICES = [
     "anushka",
@@ -371,6 +384,7 @@ TTSConfig = Annotated[
         DeepgramTTSConfiguration,
         OpenAITTSService,
         ElevenlabsTTSConfiguration,
+        CartesiaTTSConfiguration,
         DograhTTSService,
         SarvamTTSConfiguration,
     ],
@@ -485,9 +499,15 @@ class DeepgramSTTConfiguration(BaseSTTConfiguration):
     api_key: str
 
 
+CARTESIA_STT_MODELS = ["ink-whisper"]
+
+
 @register_stt
 class CartesiaSTTConfiguration(BaseSTTConfiguration):
     provider: Literal[ServiceProviders.CARTESIA] = ServiceProviders.CARTESIA
+    model: str = Field(
+        default="ink-whisper", json_schema_extra={"examples": CARTESIA_STT_MODELS}
+    )
     api_key: str
 
 
diff --git a/api/services/pipecat/service_factory.py b/api/services/pipecat/service_factory.py
index ee76263..07a7d04 100644
--- a/api/services/pipecat/service_factory.py
+++ b/api/services/pipecat/service_factory.py
@@ -7,6 +7,7 @@ from api.constants import MPS_API_URL
 from api.services.configuration.registry import ServiceProviders
 from pipecat.services.azure.llm import AzureLLMService
 from pipecat.services.cartesia.stt import CartesiaSTTService
+from pipecat.services.cartesia.tts import CartesiaTTSService
 from pipecat.services.deepgram.flux.stt import DeepgramFluxSTTService
 from pipecat.services.deepgram.stt import DeepgramSTTService, LiveOptions
 from pipecat.services.deepgram.tts import DeepgramTTSService
@@ -191,6 +192,13 @@ def create_tts_service(user_config, audio_config: "AudioConfig"):
             ),
             text_filters=[xml_function_tag_filter],
         )
+    elif user_config.tts.provider == ServiceProviders.CARTESIA.value:
+        return CartesiaTTSService(
+            api_key=user_config.tts.api_key,
+            voice_id=user_config.tts.voice,
+            model=user_config.tts.model,
+            text_filters=[xml_function_tag_filter],
+        )
     elif user_config.tts.provider == ServiceProviders.DOGRAH.value:
         # Convert HTTP URL to WebSocket URL for TTS
         base_url = MPS_API_URL.replace("http://", "ws://").replace("https://", "wss://")