Skip to main content

​
Quickstart

​
1) Sign up & create a tenant

  1. Go to LLMGrid Console: https://app.llmgrid.ai
  2. Sign up and create your Tenant (organization).
  3. Create a default Project under your tenant.
  4. Invite your team and assign roles (Admin, Developer, Viewer).
Tenants and projects enforce isolation for routes, policies, tokens, and usage.

​
2) Connect providers (BYO keys)

In Console → Providers:
  • Add OpenAI, Azure OpenAI, Anthropic, or Google Gemini keys.
  • Optionally map Azure deployments to model IDs.
  • Configure timeouts and default regions per provider.

​
3) Create a route

In Console → Routes:
  • Create chat_default route:
    • Type: Chat
    • Strategy: Cost optimized
    • Models: e.g., openai:gpt-4o-mini, anthropic:claude-3-haiku
    • Policies: rate limit, budget, retries, redaction

​
4) Generate a user token

In Console → Access:
  • Create a Project Token (scoped to your project).
  • Copy token for client usage.

​
5) First request (curl)

curl -X POST https://app.llmgrid.ai/api/v1/chat/completions \
  -H "Authorization: Bearer YOUR_PROJECT_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "tenant_id": "TENANT_123",
    "project_id": "PROJ_abc",
    "route": "chat_default",
    "messages": [
      {"role": "system", "content": "You are a helpful assistant."},
      {"role": "user", "content": "Give me 3 creative taglines for a farm supply store."}
    ],
    "stream": false
  }'

​
6) Node.js Example


import fetch from "node-fetch";

const res = await fetch("https://app.llmgrid.ai/api/v1/chat/completions", {
  method: "POST",
  headers: {
    "Authorization": "Bearer " + process.env.LLMGRID_PROJECT_TOKEN,
    "Content-Type": "application/json"
  },
  body: JSON.stringify({
    tenant_id: "TENANT_123",
    project_id: "PROJ_abc",
    route: "chat_default",
    messages: [
      { role: "system", content: "Be concise." },
      { role: "user", content: "Summarize the pros and cons of no-till farming." }
    ],
    stream: true
  })
});

// Handle streaming (SSE-like)
if (!res.ok) throw new Error(await res.text());
const reader = res.body.getReader();
for (;;) {
  const { value, done } = await reader.read();
  if (done) break;
  process.stdout.write(Buffer.from(value).toString());
}

​
6) Python Example



import requests, os

url = "https://app.llmgrid.ai/api/v1/chat/completions"
headers = {
    "Authorization": f"Bearer {os.environ['LLMGRID_PROJECT_TOKEN']}",
    "Content-Type": "application/json"
}
payload = {
    "tenant_id": "TENANT_123",
    "project_id": "PROJ_abc",
    "route": "chat_default",
    "messages": [
        {"role": "system", "content": "Answer clearly."},
        {"role": "user", "content": "What are common grain storage best practices?"}
    ],
    "stream": False
}

resp = requests.post(url, json=payload, headers=headers, timeout=60)
resp.raise_for_status()
print(resp.json())