Skip to main content

​
Overview

SlngTTSService and SlngHttpTTSService synthesize speech through SLNG, a unified voice AI gateway that routes to multiple TTS providers (Deepgram, ElevenLabs, Rime, Sarvam, and more) through a single API key. Swapping the model string switches providers with no other code changes. Use the WebSocket SlngTTSService for low-latency, interruptible streaming synthesis, or the HTTP SlngHttpTTSService for simple, non-streaming request/response synthesis.

Source Repository

Source code, examples, and issues for the SLNG integration

PyPI Package

The pipecat-slng package on PyPI

SLNG Website

Learn about the SLNG unified voice AI gateway

SLNG Docs

Documentation for the SLNG platform and APIs

​
Installation

This is a community-maintained package distributed separately from pipecat-ai: 
uv add pipecat-slng
# or
pip install pipecat-slng

​
Prerequisites

​
SLNG Account Setup

Before using the SLNG TTS services, you need:
  1. SLNG Account: Get an API key at slng.ai
  2. API Key: Used to authenticate all requests to SLNG

​
Required Environment Variables

  • SLNG_API_KEY: Your SLNG API key for authentication

​
Configuration

Both services share the same Settings dataclass (SlngTTSSettings) and most constructor parameters. The key difference is transport: SlngTTSService connects over WebSocket and exposes the full config surface, while SlngHttpTTSService issues one HTTP POST per utterance.

​
SlngTTSService (WebSocket)

​
api_key
str
required
Authentication key for the SLNG API.
​
model
str
default:"slng/deepgram/aura:2-en"
The TTS model to use. Use a slng/... prefix for SLNG-hosted routes, or an external provider route (e.g. deepgram/aura:2) proxied through SLNG.
​
voice
str
default:"None"
Voice identifier for synthesis (e.g. aura-2-thalia-en).
​
base_url
str
default:"api.slng.ai"
The API host.
​
encoding
str
default:"linear16"
Audio encoding format. One of "linear16", "mp3", "opus", "mulaw", or "alaw".
​
sample_rate
int
default:"None"
Audio sample rate in Hz. If None, uses the pipeline sample rate.
​
region_override
str
default:"None"
Pin requests to a specific datacenter. One of "ap-southeast-2", "eu-north-1", "us-east-1". Sent as the X-Region-Override header and takes precedence over world_part_override.
​
world_part_override
str
default:"None"
Constrain routing to a broad geographic zone. One of "ap", "eu", "na". Sent as the X-World-Part-Override header.
​
provider_key
str
default:"None"
Your own upstream provider API key (BYOK). Sent as the X-Slng-Provider-Key header so the provider bills your account directly. Only supported on external catalog routes (no slng/ prefix). See the BYOK docs.
​
language
Language
default:"Language.EN"
Synthesis language. Defaults to Language.EN when not given.
​
speed
float
default:"None"
Speech speed multiplier. None keeps the server default.
​
settings
SlngTTSService.Settings
default:"None"
Runtime-updatable settings override. Merged on top of any explicit kwargs.

​
SlngHttpTTSService (HTTP)

SlngHttpTTSService accepts the same parameters as SlngTTSService with these differences:
​
base_url
str
default:"https://api.slng.ai"
Full base URL (including scheme) of the SLNG API.
​
aiohttp_session
aiohttp.ClientSession
default:"None"
Optional aiohttp session. If None, one is created in start() and closed in stop()/cancel().
The HTTP bridge request body accepts only {text, voice} — there is no config object — so encoding, sample_rate, language, and speed are not configurable over HTTP. region_override/world_part_override are sent as the region/world-part query parameters. WAV responses are decoded to raw PCM; compressed responses (MP3/Ogg) yield an ErrorFrame. Use the streaming SlngTTSService if you need codec control.

​
Settings

Runtime-configurable settings passed via the settings constructor argument using SlngTTSSettings(...).
ParameterTypeDefaultDescription
voicestrNoneVoice identifier for speech synthesis.
languageLanguageLanguage.ENLanguage for speech synthesis.
speedfloatNoneSpeech speed multiplier. None uses the server default.
Changing voice, speed, or language mid-session on the WebSocket service reconnects to re-run the init handshake — expect a brief reconnect. See the source repository for the authoritative, up-to-date list.

​
Usage

​
Streaming (WebSocket)

import os
from pipecat_slng import SlngTTSService

tts = SlngTTSService(
    api_key=os.getenv("SLNG_API_KEY"),
    model="slng/deepgram/aura:2-en",
    voice="aura-2-thalia-en",
)

# ... add tts to your pipeline

​
Non-streaming (HTTP)

import os
from pipecat_slng import SlngHttpTTSService

tts = SlngHttpTTSService(
    api_key=os.getenv("SLNG_API_KEY"),
    model="slng/deepgram/aura:2-en",
    voice="aura-2-thalia-en",
)

# ... add tts to your pipeline

​
Compatibility

Tested with Pipecat v1.3.0. Check the source repository for the latest tested version and changelog.