Everything runs sovereignly on your infrastructure or via a keyed gateway. Start a 30-day Gate trial, or license a HELIX pod for your own environment.
Create an account to get a 30-day key, then drop this into Claude, Cursor, Windsurf, or any MCP client — replacing <YOUR_TRIAL_TOKEN> with your key.
{
"mcpServers": {
"verificate": {
"url": "https://mcp.verificate.ai/mcp",
"transport": "http",
"headers": { "Authorization": "Bearer <YOUR_TRIAL_TOKEN>" }
}
}
}HELIX exposes a drop-in OpenAI-compatible /v1/chat/completions endpoint with SSE streaming and a grammar-masked JSON mode — open, no API key required. It runs on any CPU (AMD, Intel, Power, Apple Silicon); licensed private deployments run entirely inside your own network.
curl https://<your-helix-pod>/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-oss-120b",
"messages": [{"role":"user","content":"Extract as JSON..."}],
"response_format": {"type": "json_object"}
}'