jargo builds real-time voice agents in Go: audio in over WebRTC, a streaming transcription → reasoning → speech pipeline with turn-taking and barge-in, and audio back out — over RTVI so existing clients interoperate.
Status:early work in progress. APIs are unstable and will change.
Pipecat is great, and jargo is a port of it — the architecture and many design decisions are Pipecat's.
This port exists for one reason: I'd rather not run a voice agent on Python.
Python is the right tool when you need the AI/data-science ecosystem. A real-time voice server doesn't: the models run as services or as ONNX, and what's left is plumbing — audio framing, WebRTC, concurrency, and shipping a binary. For that, Go is a better fit: one static binary to deploy, low and predictable memory, fast startup, and real concurrency for many simultaneous sessions without a GIL. The heavy numerics stay where they belong (the ONNX Runtime, the remote services), so giving up Python costs little here. See the benchmarks for the honest performance picture.
jargo stays on plain, standard WebRTC via Pion — no Daily, no hosted transport, no proprietary SDK or cloud to sign up for. You ship one binary, the browser connects with vanilla WebRTC, and RTVI rides the data channel. Keeping the transport open and self-hosted is a deliberate goal, not an afterthought.
WebRTC, pure Go (Pion) — audio in and out of the browser.** Opus**, not pure Go yet, waiting forpion/opusto be ready.Streaming voice pipeline: STT → LLM → TTS, with prompt caching.** Turn-taking & barge-in**: Silero VAD + Smart Turn v3, local ONNX.** RTVIdata channel — works with existing RTVI clients. Pluggable services**: swap any STT/LLM/TTS behind a small interface.** Concurrent by design**: independent processors; interruptions are frames.
jargo uses cgo (CGO_ENABLED=0
is not supported) and a few native libraries:
libsoxr— audio resampling, linked at build time (libsoxr-dev
).libopus— optional C Opus encoder, selected with-tags libopus
(libopus-dev
); the default build ships a pure-Go encoder, but libopus sounds noticeably better on speech.ONNX Runtime— loaded at run time for VAD + end-of-turn detection.
The container image bundles all of them.
go get github.com/gojargo/jargo
Locally — install the native deps, then build with cgo:
CGO_ENABLED=1 go run ./examples/echo # open http://localhost:8080
CGO_ENABLED=1 go run -tags libopus ./examples/voicebot # libopus speech encoder
With Docker — the image bundles every native dependency, so there's no host setup:
docker build -t jargo-voicebot .
docker run --rm -p 8080:8080 \
-e DEEPGRAM_API_KEY=… -e ANTHROPIC_API_KEY=… -e ELEVENLABS_API_KEY=… \
jargo-voicebot
See the ** Quickstart** for the full setup.
Two runnable bots live in examples/: an
echo bot (no API keys) and a full
voice bot (STT → LLM → TTS). The fastest way to try either — locally or with Docker — is the
.
go run ./examples/echo # then open http://localhost:8080
See ** docs/index.md** for the full documentation.
jargo is a Go port of Pipecat, distributed under the same BSD 2-Clause License. The upstream copyright — Copyright (c) 2024–2026, Daily — is preserved verbatim in LICENSE; see
for details. jargo is an independent project, not affiliated with or endorsed by Daily.
NOTICE