Skip to main content

Day 3: Running Your EchoKit Server Locally from Source Code | The First 30 Days with EchoKit

Β· 3 min read

(And Today I Finally Saw How EchoKit Works Under the Hood)

In the last two days, we powered up the EchoKit, made it talk, and connected it to a local EchoKit server using Docker. But today feels different.

Today, we’re going to build the EchoKit Server from source. No containers. No abstraction layers. Just you… and the real engine that drives EchoKit.

🌱 Why Build EchoKit from Source?​

Using Docker is great for quick setup. But building from source unlocks more possibilities:

  • Get the latest code, newest features, and bug fixes
  • Modify the server freely β€” add your own logic, prompts, or integrations
  • Compile to any platform you want
  • Truly understand how EchoKit works under the hood

If Day 1 was about getting EchoKit to speak, and Day 2 was about hosting your own server, then Day 3 is where you start becoming an EchoKit developer.

Step 1 β€” Install the Rust Toolchain​

Since the EchoKit server is written in Rust, so all the dependencies is the Rust toolchain.

Refer to the official Rust website to install Rust.

Step 2 β€” Get the Source Code​

In your terminal:

git clone https://github.com/second-state/echokit_server.git
cd echokit_server

And there it is β€” the heart of the EchoKit server, right on your machine.

It's recommended to use an IDE like VSCode to open the echokit_server folder.

πŸ”§ Step 3 β€” Configure the Server​

Open the config.toml file and fill in:

addr = "0.0.0.0:8080"
hello_wav = "hello.wav"

[tts]
platform = "Elevenlabs"
token = "sk_1234"
voice = "pNInz6obpgDQGcFmaJgB"

[asr]
url = "https://api.groq.com/openai/v1/audio/transcriptions"
api_key = "gsk_1234"
model = "whisper-large-v3"
lang = "en"
prompt = "Hello\nδ½ ε₯½\n(noise)\n(bgm)\n(silence)\n"
vad_url = "http://localhost:8000/v1/audio/vad"

[llm]
llm_chat_url = "https://api.groq.com/openai/v1/chat/completions"
api_key = "gsk_1234"
model = "openai/gpt-oss-20b"
history = 15

[[llm.sys_prompts]]
role = "system"
content = """
You are a helpful assistant. Answer truthfully and concisely. Always answer in English.
"""

Just like yesterday's setup, and remember to bring your own API key for Groq and ElevenLabs.

⚑ Step 4 β€” Build the Server​

Compile it using Cargo:

cargo build --release

The first build may take a bit, since Rust compiles everything from scratch.

But once you see:

Finished release [optimized] target(s)

you officially have your own EchoKit server binary.

πŸƒ Step 5 β€” Run the Server​

Start with debug logging:

export RUST_LOG=debug

Then launch:

nohup target/release/echokit_server &

You should see something like:

[1] 67869
appending output to nohup.out

All logs are saved in nohup.out, so you can monitor everything that happens inside your server.

πŸƒ Step 6 β€” Connect the Server​

Next, it's time to connect your EchoKit server to your device following this guide.

The EchoKit Server runs as a WebSocket service.

Your EchoKit device will connect using a URL like:

ws://YOUR_IP_ADDRESS:8080/ws

For example:

ws://192.168.1.23:8080/ws

⚠ Do NOT use localhost or 127.0.0.1. Your EchoKit device connects over Wi-Fi and cannot reach your computer’s local loopback address.

Once you enter this WebSocket URL into the EchoKit web dashboard, your device will connect to the server you built with your own hands.

Today, you didn’t just run EchoKit. You built the server. You opened the code. You compiled it. You took full control.

From now on, you’re not just a user β€” you’re a builder.

Tomorrow, in Day 4, we’ll dive into one of the most fun parts:

Customizing your EchoKit’s welcome screen. This is the beginning point where EchoKit truly becomes yours.