ElevenLabs Setup

Kartik MDT V2 integrates with ElevenLabs to generate realistic AI text-to-speech audio (used for dispatch call-outs, announcements, and other voice features in the MDT).

All ElevenLabs credentials are configured in:

kartik-mdt-v2/shared/server_config.lua

[!IMPORTANT] server_config.lua is server-side only. Never expose these keys in shared/config.lua or anywhere reachable by the client.

🔧 Configuration BlockCopy Anchor Link

Open shared/server_config.lua and locate the ServerConfig.ElevenLabs table:

ServerConfig.ElevenLabs = {
    enabled = false,
    apiKey = "",
    voiceId = "",
    modelId = "",
    stability = 0.5,
    similarityBoost = 0.5
}

🔑 1. Where to Get the API KeyCopy Anchor Link

1. Go to https://elevenlabs.io/ and sign in (or create a free account).
2. Click your profile icon (top right) → Profile + API Key (or My Account → API Keys).
3. Click Create API Key, give it a name (e.g. FiveM MDT), and copy the key.
4. Paste it into apiKey = "".

apiKey = "sk_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",

[!WARNING] Treat the API key like a password. Anyone with it can spend your ElevenLabs credits.

🗣️ 2. Where to Get the Voice IDCopy Anchor Link

1. In the ElevenLabs dashboard, open the Voices section (left sidebar → Voices / VoiceLab).
2. Pick a voice from the Voice Library (add it to My Voices) or use one you've cloned.
3. Click the voice → open the ⋮ (three dots) menu → Copy Voice ID.
   • Alternatively, the Voice ID is shown in the URL when viewing the voice.
4. Paste it into voiceId = "".

voiceId = "21m00Tcm4TlvDq8ikWAM",

🤖 3. Where to Get the Model IDCopy Anchor Link

ElevenLabs offers multiple models. Common values:

To see the full list:

1. Visit https://elevenlabs.io/docs/api-reference/models, or
2. In the dashboard, go to Speech Synthesis, the model dropdown lists every available model_id.

modelId = "eleven_multilingual_v2",

🎛️ 4. Stability & Similarity BoostCopy Anchor Link

These two sliders control how the voice sounds. You can preview them live in the ElevenLabs Speech Synthesis playground before committing values.

• stability (default 0.5)
  • Low (≈0.2) → more emotion and variation, but can sound unstable.
  • High (≈0.75) → very consistent, but flatter delivery.
• similarityBoost (default 0.5)
  • Low → more creative, can drift from the original voice.
  • High → sticks closely to the voice sample.

For police/dispatch style call-outs, stability = 0.6 and similarityBoost = 0.75 is a good starting point.

✅ Final ExampleCopy Anchor Link

ServerConfig.ElevenLabs = {
    enabled = true,
    apiKey = "sk_xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx",
    voiceId = "21m00Tcm4TlvDq8ikWAM",
    modelId = "eleven_multilingual_v2",
    stability = 0.6,
    similarityBoost = 0.75
}

After saving the file, restart the kartik-mdt-v2 resource:

ensure kartik-mdt-v2

or in the live console:

restart kartik-mdt-v2