llm

Multi-provider LLM API integration for AI agent workflows.

use llm

Functions

Function	Signature	Description
chat	(provider, model, messages, options?) -> Result	Send chat completion request
complete	(provider, model, prompt, options?) -> Result	Simple text completion
stream	(provider, model, messages, callback) -> Result	Streaming chat (calls callback with response)
models	(provider) -> [String]	List available models
estimate_cost	(model, input_tokens, output_tokens) -> Cost	Estimate API cost in USD
providers	() -> [Provider]	List supported providers

Supported providers: "openai", "anthropic"

Example

use llm

let result = llm.chat("openai", "gpt-4o", [
  {role: "system", content: "You are a helpful assistant."},
  {role: "user", content: "What is Arc?"}
], {temperature: 0.7})

if result.ok {
  print(result.content)
}

# Estimate costs
let cost = llm.estimate_cost("gpt-4o", 1000, 500)
print("Cost: ${cost.total}")