LLM Providers

Overview

Fiberwise provides a unified interface for working with multiple Large Language Model (LLM) providers. This abstraction layer allows you to switch between providers, compare outputs, and leverage the best features of each platform without changing your code.

Provider Management

Add and manage LLM providers using the Fiberwise CLI:

🔧 Adding Providers

# Add OpenAI provider
fiber account add-provider --provider openai --api-key "sk-..." --model gpt-4 --set-default

# Add Anthropic provider
fiber account add-provider --provider anthropic --api-key "sk-ant-..." --model claude-3-5-sonnet-20241022 --set-default

# Add Google provider
fiber account add-provider --provider google --api-key "AIza..." --model gemini-pro

# Add Hugging Face provider (free tier available)
fiber account add-provider --provider huggingface --api-key "hf_..." --model "meta-llama/Llama-2-7b-chat-hf"

# Add OpenRouter provider (with free model)
fiber account add-provider --provider openrouter --api-key "sk-or-..." --model "meta-llama/llama-3.1-8b-instruct:free" --site-url "https://yourapp.com"

# Add Cloudflare Workers AI provider
fiber account add-provider --provider cloudflare --api-key "xxx" --account-id "xxx" --model "@cf/meta/llama-3.1-8b-instruct"

📋 Listing Providers

# View all configured providers
fiber account list-providers

# View provider configurations
fiber account list-configs

Key Features

API Usage

Use LLM providers in your agents and functions:

🐍 Python SDK

from fiberwise_sdk import FiberAgent, LLMProvider

class MyAgent(FiberAgent):
    def __init__(self):
        super().__init__()
        # Provider is automatically injected
        self.llm = self.get_llm_provider()

    def process(self, input_data):
        # Use the configured default provider
        response = self.llm.complete(
            prompt=f"Analyze this data: {input_data}",
            model="gpt-4"  # Optional: override default
        )
        return {"analysis": response}

🌐 REST API

curl -X POST http://localhost:7001/api/v1/llm-providers/complete \
  -H "Authorization: Bearer your-api-key" \
  -H "Content-Type: application/json" \
  -d '{
    "provider": "openai",
    "model": "gpt-4",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Configuration

Provider configuration options and environment variables:

⚙️ Environment Variables

# Provider API Keys
export OPENAI_API_KEY="sk-your-key-here"
export ANTHROPIC_API_KEY="sk-ant-your-key-here"
export GOOGLE_API_KEY="AIza-your-key-here"

# Default Provider
export FIBERWISE_DEFAULT_LLM_PROVIDER="anthropic"
export FIBERWISE_DEFAULT_MODEL="claude-3-sonnet"

📄 Configuration File

# fiberwise.yaml
llm_providers:
  default: "anthropic"
  providers:
    openai:
      api_key: "${OPENAI_API_KEY}"
      models: ["gpt-4", "gpt-3.5-turbo"]
    anthropic:
      api_key: "${ANTHROPIC_API_KEY}"
      models: ["claude-3-sonnet", "claude-3-haiku"]

Best Practices