Documentation

Get started in minutes.

Forge is a drop-in replacement. If you've used the OpenAI or Anthropic API, you already know how to use Forge.

1

Get your API key

Sign up and generate an API key from your dashboard. Keys start with rrt-burst-

2

Point your SDK at Forge

Python (OpenAI SDK) 
# pip install openai
from openai import OpenAI

client = OpenAI(
    base_url="https://forge.lanaai.io/v1",
    api_key="your-forge-api-key",
)

response = client.chat.completions.create(
    model="auto",  # or "fast", "gpt-4o", "claude-sonnet-4-20250514", etc.
    messages=[{"role": "user", "content": "Hello from Forge!"}],
)

print(response.choices[0].message.content)
cURL 
curl https://forge.lanaai.io/v1/chat/completions \
  -H "Authorization: Bearer $FORGE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "auto",
    "messages": [{"role": "user", "content": "Hello"}],
    "stream": true
  }'
Anthropic SDK 
# pip install anthropic
import anthropic

client = anthropic.Anthropic(
    base_url="https://forge.lanaai.io/v1",
    api_key="your-forge-api-key",
)

message = client.messages.create(
    model="claude-sonnet-4-20250514",
    max_tokens=1024,
    messages=[{"role": "user", "content": "Hello from Forge!"}],
)
3

That's it.

Forge handles routing, cost optimization, and failover automatically. Monitor your usage from the dashboard.

API Endpoints

POST
/v1/chat/completions

OpenAI-compatible chat completions. Supports streaming.

POST
/v1/messages

Anthropic Messages API compatible. Full streaming SSE support.

POST
/v1/embeddings

Generate embeddings with the same interface as OpenAI.

GET
/v1/models

List available models and their capabilities.

GET
/dashboard/usage

Your token usage, cost breakdown, and forecasted spend.

Model Aliases

Let Forge pick the best model for you, or request a specific one.

Alias Routes to Best for
"auto" Cost-optimized pick General use — cheapest that works
"fast" Smallest, fastest model Low latency, simple tasks
"reasoning" Strongest available model Complex analysis, coding, research
"embedding" Embedding model Vector embeddings for RAG

You can also use specific model names like "gpt-4o", "claude-sonnet-4-20250514", "llama-3.1-70b", etc.