SuperconGPT¶

3 tabs: (1) Inverse design — generate crystal structures for target Tc using AtomGPT, (2) Database search for known superconductors, (3) Predict Tc for any structure with ALIGNN.

Open App

Overview¶

3 tabs: (1) Inverse design — generate crystal structures for target Tc using AtomGPT, (2) Database search for known superconductors, (3) Predict Tc for any structure with ALIGNN.

Data Source

dft_3d + AtomGPT + ALIGNN supercon model

Endpoints¶

GET /supercon
POST /supercon/generate
GET /supercon/generate
GET /supercon/search
POST /supercon/predict_tc

Request Models: SuperconGenerateRequest, SuperconPredictRequest

Authentication

All POST endpoints require Authorization: Bearer YOUR_TOKEN.

API Example¶

import requests

response = requests.post(
    "https://atomgpt.org/supercon/generate",
    headers={
        "Authorization": "Bearer sk-XYZ",
        "accept": "application/json",
        "Content-Type": "application/json",
    },
    json={"jid": "JVASP-1002"},
)
data = response.json()
print(data)

AGAPI Agent¶

from agapi.agents import AGAPIAgent
import os

agent = AGAPIAgent(api_key=os.environ.get("AGAPI_KEY"))
response = agent.query_sync("Show supercongpt for Silicon")
print(response)

Reference¶

NPJ Comp. Mat. 8, 244 (2023); J. Phys. Chem. Lett. 15, 6909 (2024)