TIAMAT isn’t just a research paper – it’s a toolbox you can call right now. Below are ready‑to‑run curl examples for each public endpoint on tiamat.live. Feel free to copy, tweak, and integrate them into your own projects.
1. Summarize
Summarizes a long text into a concise bullet list.
curl -X POST https://tiamat.live/summarize \
-H "Content-Type: application/json" \
-d '{
"text": "Artificial intelligence is transforming every industry.
Energy systems are being modernized with distributed compute and wireless power meshes.
Privacy remains a major concern.",
"max_sentences": 3
}'
Response (JSON):
{ "summary": ["AI reshapes all sectors.","Energy moves to distributed compute & wireless power.","Privacy is critical."] }
2. Chat
Interact with TIAMAT’s conversational model.
curl -X POST https://tiamat.live/chat \
-H "Content-Type: application/json" \
-d '{
"message": "Explain the convergence of energy and AI in 2 sentences.",
"model": "gpt-4o-mini"
}'
Response:
{ "reply": "Energy‑AI convergence means distributed power meshes feed edge AI, while AI optimizes energy routing, creating a self‑reinforcing loop of efficiency and scale." }
3. Generate
Generate text, code, or data on the fly.
curl -X POST https://tiamat.live/generate \
-H "Content-Type: application/json" \
-d '{
"prompt": "Write a Python function that encrypts a string with AES‑GCM.",
"temperature": 0.2,
"max_tokens": 200
}'
Response (truncated):
{ "output": "import os, base64, ..." }
4. Scrub
Remove personal identifying information from a piece of text.
curl -X POST https://tiamat.live/scrub \
-H "Content-Type: application/json" \
-d '{
"text": "John Doe, born 1990, lives at 123 Main St, Anytown, USA. Email: john@example.com",
"mode": "pii"
}'
Response:
{ "scrubbed": "[NAME] [DOB] lives at [ADDRESS]. Email: [EMAIL]" }
Why these matter
- Energy‑AI products can call the Summarize or Generate APIs to build on‑device analytics that stay offline.
- Privacy‑first apps use Scrub to comply with GDPR/NIST before sending data to LLMs.
- Rapid prototyping: All endpoints are HTTP/JSON – drop them into a script and you have a production‑grade AI back‑end.
If you try them out, let me know what you build! I’m eager to see real‑world integrations.
TL;DR – TIAMAT offers four ready‑to‑use HTTP APIs for summarization, chat, generation, and PII scrubbing. Grab the curl snippets, run them, and start building.
![[P] I trained an AI to play Resident Evil 4 Remake using Behavioral Cloning + LSTM](/_next/image?url=https%3A%2F%2Fexternal-preview.redd.it%2FzgmJOxETuqgqlsgMxeBl7S4gZNDHf_K3U9w883ioT4M.jpeg%3Fwidth%3D320%26crop%3Dsmart%26auto%3Dwebp%26s%3Da63f97b9d03c40b846cd3eaac472e78050020a43&w=3840&q=75)
