curllm

LLM Providers Examples

Examples showing how to use curllm with different LLM providers.

Supported Providers

curllm supports multiple LLM providers via litellm:

Provider	Format	Environment Variable
Ollama (local)	ollama/qwen2.5:7b	-
OpenAI	openai/gpt-4o-mini	OPENAI_API_KEY
Anthropic	anthropic/claude-3-haiku-20240307	ANTHROPIC_API_KEY
Gemini	gemini/gemini-2.0-flash	GEMINI_API_KEY
Groq	groq/llama3-70b-8192	GROQ_API_KEY
DeepSeek	deepseek/deepseek-chat	DEEPSEEK_API_KEY

Examples

Basic Usage

from curllm_core import CurllmExecutor, LLMConfig

# Default: local Ollama
executor = CurllmExecutor()

# OpenAI (auto-detects OPENAI_API_KEY)
executor = CurllmExecutor(LLMConfig(provider="openai/gpt-4o-mini"))

# Explicit API token
executor = CurllmExecutor(LLMConfig(
    provider="anthropic/claude-3-haiku-20240307",
    api_token="sk-ant-..."
))

Run Examples

# Set your API key first
export OPENAI_API_KEY="sk-..."

# Run example
python examples/llm-providers/openai_example.py
python examples/llm-providers/multi_provider_benchmark.py

Files

• openai_example.py - OpenAI GPT-4o usage
• anthropic_example.py - Anthropic Claude usage
• gemini_example.py - Google Gemini usage
• groq_example.py - Groq (fast Llama) usage
• multi_provider_benchmark.py - Compare providers
• auto_detect_provider.py - Auto-detect from environment

Related

• LLMConfig Documentation
• Main Examples

This site is open source. Improve this page.