Connect Together AI to ElevenLabs Conversational AI Voice Agents

• High-Performance Inference
  • Sub-second model processing with research-driven optimizations
  • Custom kernels, quantization, and speculative decoding for maximum efficiency
  • Consistent low-latency responses even during peak traffic loads
• Model Flexibility & Choice
  • Access to extensive model library with state-of-the-art options
  • Deploy lightweight models for speed or massive models for complex dialogues
  • Easy model swapping without code changes—test and deploy in minutes
• Custom Model Deployment
  • Bring your own fine-tuned models trained on proprietary data
  • Deploy models that speak your company's language and domain knowledge
  • Maintain full IP control while leveraging enterprise-grade infrastructure
• Developer-First Integration
  • Standard OpenAI-compatible API eliminates vendor lock-in
  • Plug-and-play setup with your existing ElevenLabs agents
  • Comprehensive API documentation and enterprise support
• Enterprise Security & Compliance
  • SOC 2 & HIPAA compliant infrastructure with GDPR alignment
  • Data encrypted in transit with zero retention mode available
  • Enterprise SLAs with dedicated support channels

Part 1: Together AI Setup

Obtain API Key: Navigate to Together AI Settings → API Keys, create new key
Select Model: Choose from catalog or use your custom fine-tuned model
Note Model ID: Copy exact identifier (e.g., "meta-llama/Meta-Llama-3.1-70B-Instruct-Turbo")

Part 2: ElevenLabs Setup

In agent settings, add Together AI key as "Custom LLM" secret

Part 3 Configure Custom LLM:

Select Custom LLM from the LLM dropdown and save the following details:

Server URL: https://api.together.xyz/v1
Model ID: Your chosen model identifier
Credentials: Select the API key from dropdown

Save the details and use "Test AI Agent" to verify integration