Your first gateway request

From zero to a working LLM call through the 1KeyCore gateway in under 2 minutes. You need a 1KeyCore account and at least one provider API key added to your workspace.

Create an account & workspace

Sign up at 1keycore.saurabh.codes/signup. After signup you'll go through a short onboarding — either create a new workspace (you become the admin) or join an existing one using an invite code from your admin. You'll land on the dashboard once that's done.

Add a provider key (admins)

If you're the workspace admin, go to Dashboard → API Keys, select a provider (OpenAI, Anthropic, Gemini…), paste your provider key, and click Save key. The raw key is AES-256 encrypted on write — we never log or expose it.

Once saved, each key gets a Request Key — a short token used as the x-api-key header in every chat request. Click copy next to it in the keys table.

If you joined a workspace, the admin's keys are already available — head to Dashboard → API Keys to copy the request key for the provider you want to use.

Make your first request

Send a POST to /chat with your Request Key in the x-api-key header. The gateway resolves the provider and model automatically.

shell

curl -X POST https://1keyapi.saurabh.codes/chat \
  -H "x-api-key: <your_request_key>" \
  -H "Content-Type: application/json" \
  -d '{
    "message": "Summarise this contract in 3 bullets"
    "model":   "gpt-4o",
    "temperature":"0"
    "max_token":"2000",
    "best_model_choice":"true",
    "system_prompt": "string"
  }'

response · 200 OK

{
  "response":   "1. The contract grants...",
  "model":      "gpt-4o",
  "tokens":     312,
  "latency_ms": 38,
  "cached":     false
}

The same request key works for any provider key in your workspace — swap the model field to route to a different provider.

request fields

best_model_choice

Enables intelligent routing. When true, the gateway automatically selects the best model for the task based on performance, cost, and latency.

model

Specifies the model explicitly. Overrides intelligent routing. Only a small set of supported models are available currently (e.g. gpt-4o, llama-3.3-70b-versatile, openai/gpt-oss-120b).

temperature

Controls randomness. Lower values (e.g. 0) produce more deterministic responses, while higher values increase creativity.

Invite your team

As a workspace admin, go to Dashboard click on Invite member and generate an invite code. Share it with your teammates — they enter it during onboarding to join your workspace and immediately get access to the keys you've added.

Each member's requests are tracked separately in usage, so you can see per-user consumption without sharing credentials.

next steps

← Back to docs Go to dashboard →Manage keys →