open-notebook/docs/5-CONFIGURATION/local-tts.md

Local Text-to-Speech Setup

Run text-to-speech locally for free, private podcast generation using OpenAI-compatible TTS servers.

---

Why Local TTS?

Benefit	Description
Free	No per-character costs after setup
Private	Audio never leaves your machine
Unlimited	No rate limits or quotas
Offline	Works without internet

---

Quick Start with Speaches

Speaches is an open-source, OpenAI-compatible TTS server.

Step 1: Create Docker Compose File

Create a folder and add docker-compose.yml:

services:
  speaches:
    image: ghcr.io/speaches-ai/speaches:latest-cpu
    container_name: speaches
    ports:
      - "8969:8000"
    volumes:
      - hf-hub-cache:/home/ubuntu/.cache/huggingface/hub
    restart: unless-stopped

volumes:
  hf-hub-cache:

Step 2: Start and Download Model

# Start Speaches
docker compose up -d

# Wait for startup
sleep 10

# Download voice model (~500MB)
docker compose exec speaches uv tool run speaches-cli model download speaches-ai/Kokoro-82M-v1.0-ONNX

Step 3: Test

curl "http://localhost:8969/v1/audio/speech" -s \
  -H "Content-Type: application/json" \
  --output test.mp3 \
  --data '{
    "input": "Hello! Local TTS is working.",
    "model": "speaches-ai/Kokoro-82M-v1.0-ONNX",
    "voice": "af_bella"
  }'

Play test.mp3 to verify.

Step 4: Configure Open Notebook

Docker deployment:

# In your Open Notebook docker-compose.yml
environment:
  - OPENAI_COMPATIBLE_BASE_URL_TTS=http://host.docker.internal:8969/v1

Local development:

export OPENAI_COMPATIBLE_BASE_URL_TTS=http://localhost:8969/v1

Step 5: Add Model in Open Notebook

1. Go to Settings → Models
2. Click Add Model in Text-to-Speech section
3. Configure:
   • Provider: openai_compatible
   • Model Name: speaches-ai/Kokoro-82M-v1.0-ONNX
   • Display Name: Local TTS
4. Click Save
5. Set as default if desired

---

Available Voices

The Kokoro model includes multiple voices:

Female Voices

Voice ID	Description
af_bella	Clear, professional
af_sarah	Warm, friendly
af_nicole	Energetic, expressive

Male Voices

Voice ID	Description
am_adam	Deep, authoritative
am_michael	Friendly, conversational

British Accents

Voice ID	Description
bf_emma	British female, professional
bm_george	British male, formal

Test Different Voices

for voice in af_bella af_sarah am_adam am_michael; do
  curl "http://localhost:8969/v1/audio/speech" -s \
    -H "Content-Type: application/json" \
    --output "test_${voice}.mp3" \
    --data "{
      \"input\": \"Hello, this is the ${voice} voice.\",
      \"model\": \"speaches-ai/Kokoro-82M-v1.0-ONNX\",
      \"voice\": \"${voice}\"
    }"
done

---

GPU Acceleration

For faster generation with NVIDIA GPUs:

services:
  speaches:
    image: ghcr.io/speaches-ai/speaches:latest-cuda
    container_name: speaches
    ports:
      - "8969:8000"
    volumes:
      - hf-hub-cache:/home/ubuntu/.cache/huggingface/hub
    restart: unless-stopped
    deploy:
      resources:
        reservations:
          devices:
            - driver: nvidia
              count: 1
              capabilities: [gpu]

volumes:
  hf-hub-cache:

---

Docker Networking

Open Notebook in Docker (macOS/Windows)

OPENAI_COMPATIBLE_BASE_URL_TTS=http://host.docker.internal:8969/v1

Open Notebook in Docker (Linux)

# Option 1: Docker bridge IP
OPENAI_COMPATIBLE_BASE_URL_TTS=http://172.17.0.1:8969/v1

# Option 2: Host networking
docker run --network host ...

Remote Server

Run Speaches on a different machine:

# On server, bind to all interfaces
# Then in Open Notebook:
OPENAI_COMPATIBLE_BASE_URL_TTS=http://server-ip:8969/v1

---

Multi-Speaker Podcasts

Configure different voices for each speaker:

Speaker 1 (Host):
  Model: speaches-ai/Kokoro-82M-v1.0-ONNX
  Voice: af_bella

Speaker 2 (Guest):
  Model: speaches-ai/Kokoro-82M-v1.0-ONNX
  Voice: am_adam

Speaker 3 (Narrator):
  Model: speaches-ai/Kokoro-82M-v1.0-ONNX
  Voice: bf_emma

---

Troubleshooting

Service Won't Start

# Check logs
docker compose logs speaches

# Verify port available
lsof -i :8969

# Restart
docker compose down && docker compose up -d

Connection Refused

# Test Speaches is running
curl http://localhost:8969/v1/models

# From inside Open Notebook container
docker exec -it open-notebook curl http://host.docker.internal:8969/v1/models

Model Not Found

# List downloaded models
docker compose exec speaches uv tool run speaches-cli model list

# Download if missing
docker compose exec speaches uv tool run speaches-cli model download speaches-ai/Kokoro-82M-v1.0-ONNX

Poor Audio Quality

• Try different voices
• Adjust speed: "speed": 0.9 to 1.2
• Check model downloaded completely
• Allocate more memory

Slow Generation

Solution	How
Use GPU	Switch to latest-cuda image
More CPU	Allocate more cores in Docker
Faster model	Use smaller/quantized models
SSD storage	Move Docker volumes to SSD

---

Performance Tips

Recommended Specs

Component	Minimum	Recommended
CPU	2 cores	4+ cores
RAM	2 GB	4+ GB
Storage	5 GB	10 GB (for multiple models)
GPU	None	NVIDIA (optional)

Resource Limits

services:
  speaches:
    # ... other config
    mem_limit: 4g
    cpus: 2

Monitor Usage

docker stats speaches

---

Comparison: Local vs Cloud

Aspect	Local (Speaches)	Cloud (OpenAI/ElevenLabs)
Cost	Free	$0.015-0.10/min
Privacy	Complete	Data sent to provider
Speed	Depends on hardware	Usually faster
Quality	Good	Excellent
Setup	Moderate	Simple API key
Offline	Yes	No
Voices	Limited	Many options

When to Use Local

• Privacy-sensitive content
• High-volume generation
• Development/testing
• Offline environments
• Cost control

When to Use Cloud

• Premium quality needs
• Multiple languages
• Time-sensitive projects
• Limited hardware

---

Other Local TTS Options

Any OpenAI-compatible TTS server works. The key is:

1. Server implements /v1/audio/speech endpoint
2. Set OPENAI_COMPATIBLE_BASE_URL_TTS to server URL
3. Add model with provider openai_compatible

---

Related

• OpenAI-Compatible Providers - General compatible provider setup
• AI Providers - All provider configuration
• Creating Podcasts - Using TTS for podcasts