didlogic is now integrated with Pipecat

We are pleased to announce that didlogic is now part of the Pipecat documentation, listed as a community integration under frame serializers.

Pipecat is the open-source Python framework many teams use to build real-time voice agents. With the new didlogic package for Pipecat, a bot can answer calls to a didlogic number and place outbound calls through our API, with no SIP stack to run.

What the integration is

The integration is a serializer: a small adapter that translates call audio between Pipecat and the didlogic voice-streaming protocol.

It plugs into Pipecat’s standard WebSocket transport. Everything else in the pipeline (speech-to-text, LLM, text-to-speech) stays ordinary Pipecat, so you keep full freedom over the models and vendors you use.

The package is maintained by didlogic and published on PyPI, with source code and examples on GitHub.

What it enables

• Inbound calls. Point a didlogic number at your bot’s endpoint and every call to that number is streamed straight to it.
• Outbound calls. Place calls through the didlogic Click2Call API, with your bot as the agent on the line.
• No SIP trunk to manage. didlogic connects to your server over WebSocket. There is no SIP server, registration or codec negotiation on your side.
• Natural conversations. Interruptions are handled by clearing buffered audio, so the bot stops talking the moment the caller does.
• Real call control. The bot knows when the person actually picks up, keypad presses (DTMF) reach the pipeline, and ending the conversation hangs up the call.
• Global reach. Your bot can use local and toll-free numbers in 140+ countries.

How it works

didlogic connects to your server, so the bot acts as the WebSocket server. One protocol carries both directions of call.

Audio travels as PCM16 at 24 kHz mono, in 10 ms frames. That is a wideband format, so speech recognition gets cleaner input than classic 8 kHz telephony audio.

On the developer side, the didlogic-specific part of a bot is small: one package and a couple of transport settings.

Our Pipecat setup guide walks through a complete, ready-to-run server for both inbound and outbound calls, starting from the Pipecat quickstart.

What you need

• A didlogic account with voice streaming enabled. If the WebSocket destination is not visible in your portal, our support team can turn it on.
• A didlogic number for inbound calls, or an API token and SIP account for outbound calls.
• A publicly reachable secure WebSocket address (a tunnel such as ngrok works for testing).
• API keys for the speech and LLM services you use in your pipeline.

Part of a growing list of AI platforms

Pipecat joins ElevenLabs, LiveKit, Retell and Vapi in our AI gateways documentation. It is the first of these connected over WebSocket rather than SIP, which makes it the simplest path for teams building their own voice stack in Python.

Getting started

• Pipecat docs: didlogic frame serializer
• didlogic: Pipecat WebSocket calls guide
• pipecat-didlogic on GitHub
• Open a didlogic account

New accounts include free test credit, so you can run your first call before committing to anything.

We are glad to see didlogic in the Pipecat documentation, and we look forward to carrying the calls for the voice agents our customers build with it.