Brick-v1-beta

Brick-v1-beta is a lightweight prompt-complexity classifier designed for LLM routing pipelines. It scores each incoming prompt as easy, medium, or hard, helping a router send requests to the most cost-efficient model tier. The eco variant is specifically optimized for cost control, preferring the cheaper path when uncertainty is high.

Getting Started

pip install requests

import requests

api_url = "https://api.regolo.ai/v1/chat/completions"
headers = {
    "Content-Type": "application/json",
    "Authorization": "Bearer YOUR_REGOLO_KEY"
}
data = {
  "model": "brick-v1-beta",
  "messages": [
    {
      "role": "user",
      "content": "What is the capital of Italy, and which region does it belong to?"
    }
  ]
}

response = requests.post(api_url, headers=headers, json=data)
print(response.json())Code language: Bash (bash)

Output

{
  "id": "chatcmpl-a4988541-84b1-41a5-843f-06790a11f7fc",
  "created": 1769560420,
  "model": "hosted_vllm/brick-v1-beta",
  "object": "chat.completion",
  "choices": [
    {
      "finish_reason": "stop",
      "index": 0,
      "message": {
        "content": "The capital of Italy is Rome (Italian: Roma). Rome belongs to the Lazio region.",
        "role": "assistant"
      }
    }
  ],
  "usage": {
    "completion_tokens": 62,
    "prompt_tokens": 45,
    "total_tokens": 107
  }
}Code language: JSON / JSON with Comments (json)

Additional Information

Applications & Use Cases

• LLM routing systems that need a fast complexity estimate before generation.
• Cost-aware API gateways that decide when a request can stay on a cheaper model tier.
• Multi-model assistants that reserve stronger models for harder prompts.
• Enterprise AI orchestration flows focused on lowering inference spend at scale.
• High-volume workloads where prompt difficulty helps determine the right serving tier.