XRD / DiffractGPT¶

XRD analysis suite: simulate powder XRD patterns from crystal structures, match experimental data to JARVIS-DFT and COD, Rietveld-style refinement, AI-powered structure prediction via DiffractGPT, and optional ALIGNN/SlakoNet property predictions on structures.

Open App

Overview¶

The XRD app provides a full pipeline from experimental diffraction data to crystal structure identification and property prediction. Three analysis methods are available: cosine-similarity pattern matching against JARVIS-DFT + COD databases, AI-powered structure generation via DiffractGPT, and Rietveld refinement via DARA. Results include best-match POSCAR, similarity scores, overlay plots, and optional ALIGNN/SlakoNet predictions.

Data Source

dft_3d (76K materials) + COD (Crystallography Open Database) for pattern matching. DiffractGPT model for AI structure prediction from peaks.

Endpoints¶

`POST /xrd/analyze` — Match XRD pattern to structures¶

curl -X POST "https://atomgpt.org/xrd/analyze" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{
    "formula": "Si",
    "xrd_data": "28.44 1.00\n47.30 0.55\n56.12 0.30\n69.13 0.06\n76.38 0.11",
    "wavelength": 1.54184,
    "method": "pattern_matching",
    "interval": 0.1,
    "x_range_min": 0.0,
    "x_range_max": 90.0
  }'

Options for method: "pattern_matching" (default), "diffractgpt", or "both".

Use DiffractGPT:

curl -X POST "https://atomgpt.org/xrd/analyze" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{
    "formula": "LaB6",
    "xrd_data": "21.36 1.00\n30.39 0.64\n37.44 0.31\n43.51 0.20",
    "method": "diffractgpt"
  }'

Run both methods:

curl -X POST "https://atomgpt.org/xrd/analyze" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{
    "formula": "SrTiO3",
    "xrd_data": "22.75 0.30\n32.40 1.00\n39.95 0.45\n46.47 0.60\n57.79 0.25",
    "method": "both"
  }'

Field	Type	Default	Description
`formula`	string	required	Chemical formula (e.g. `"Si"`, `"LaB6"`, `"SrTiO3"`)
`xrd_data`	string	required	`2theta intensity` pairs, newline-separated
`wavelength`	float	1.54184	X-ray wavelength in Å (Cu Kα)
`method`	string	`"pattern_matching"`	`"pattern_matching"`, `"diffractgpt"`, or `"both"`
`interval`	float	0.1	2θ interpolation step (degrees)
`x_range_min`	float	0.0	Minimum 2θ range
`x_range_max`	float	90.0	Maximum 2θ range

`POST /xrd/analyze_with_refinement` — Full pipeline with Rietveld + ALIGNN¶

curl -X POST "https://atomgpt.org/xrd/analyze_with_refinement" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{
    "formula": "Si",
    "xrd_data": "28.44 1.00\n47.30 0.55\n56.12 0.30",
    "wavelength": 1.54184,
    "method": "both",
    "run_refinement": true,
    "run_alignn": true,
    "run_slakonet": false
  }'

Field	Type	Default	Description
`formula`	string	required	Chemical formula
`xrd_data`	string	required	2θ/intensity data
`wavelength`	float	1.54184	X-ray wavelength (Å)
`method`	string	`"pattern_matching"`	`"pattern_matching"`, `"diffractgpt"`, or `"both"`
`run_refinement`	bool	`true`	Run Rietveld refinement on best match
`run_alignn`	bool	`false`	Predict properties with ALIGNN on matched structure
`run_slakonet`	bool	`false`	Compute band structure with SlakoNet
`best_match_poscar`	string	`""`	Override starting structure for refinement

Pipeline steps: (1) Pattern matching → (2) DiffractGPT → (3) Rietveld refinement → (4) ALIGNN/SlakoNet predictions.

`POST /xrd/generate` — Generate XRD from POSCAR¶

curl -X POST "https://atomgpt.org/xrd/generate" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{
    "poscar": "Si\n1.0\n0 2.734 2.734\n2.734 0 2.734\n2.734 2.734 0\nSi\n2\ndirect\n0 0 0\n0.25 0.25 0.25",
    "wavelength": 1.54184
  }'

Returns two_thetas, d_hkls, intensities (normalized to 100), formula, spacegroup, lattice parameters.

`GET /diffractgpt/query` — Direct DiffractGPT prediction¶

curl "https://atomgpt.org/diffractgpt/query?formula=Si&peaks=28.4(1.0),47.3(0.55),56.1(0.30)&max_new_tokens=456" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "accept: application/json"

Param	Description
`formula`	Chemical formula
`peaks`	Peaks as `2theta(intensity)` comma-separated
`max_new_tokens`	Max generation tokens (default 456)

Returns plain-text POSCAR of the predicted structure.

`GET /pxrd/query` — Legacy PXRD pattern matching¶

curl "https://atomgpt.org/pxrd/query?pattern=Si%0A28.44%201.00%0A47.30%200.55%0A56.12%200.30&intvl=0.1&x_min=0&x_max=90" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "accept: application/json"

Pattern format: first line = formula, subsequent lines = 2theta intensity. URL-encode newlines as %0A.

`POST /xrd/poscar_to_xyz` — Convert POSCAR to XYZ¶

curl -X POST "https://atomgpt.org/xrd/poscar_to_xyz" \
  -H "Authorization: Bearer sk-XYZ" \
  -H "Content-Type: application/json" \
  -H "accept: application/json" \
  -d '{"poscar": "Si\n1.0\n0 2.734 2.734\n2.734 0 2.734\n2.734 2.734 0\nSi\n2\ndirect\n0 0 0\n0.25 0.25 0.25"}'

Python Example¶

import requests

response = requests.post(
    "https://atomgpt.org/xrd/analyze",
    headers={
        "Authorization": "Bearer sk-XYZ",
        "accept": "application/json",
        "Content-Type": "application/json",
    },
    json={
        "formula": "Si",
        "xrd_data": "28.44 1.00\n47.30 0.55\n56.12 0.30",
        "method": "both",
    },
)
data = response.json()

# Pattern matching result
pm = data.get("pattern_matching", {})
if pm.get("success"):
    best = pm["best_match"]
    print(f"Best match: {best['jid']} ({best['formula']})")
    print(f"Similarity: {best['similarity']*100:.1f}%")
    print(f"POSCAR:\n{best['poscar'][:200]}")

# DiffractGPT result
dg = data.get("diffractgpt", {})
if dg.get("success"):
    s = dg["structure"]
    print(f"DiffractGPT: {s['formula']} ({s['spacegroup']})")
    print(f"Similarity: {s['similarity']*100:.1f}%")

AGAPI Agent¶

from agapi.agents import AGAPIAgent
import os

agent = AGAPIAgent(api_key=os.environ.get("AGAPI_KEY"))
response = agent.query_sync("Match this XRD pattern for Silicon: 28.44 1.00, 47.30 0.55, 56.12 0.30")
print(response)

Reference¶

J. Phys. Chem. Lett. 16, 2110 (2025) — DiffractGPT DOI
atomgptlab/jarvis
atomgptlab/agapi

XRD / DiffractGPT¶

Overview¶

Endpoints¶

POST /xrd/analyze — Match XRD pattern to structures¶

POST /xrd/analyze_with_refinement — Full pipeline with Rietveld + ALIGNN¶

POST /xrd/generate — Generate XRD from POSCAR¶

GET /diffractgpt/query — Direct DiffractGPT prediction¶

GET /pxrd/query — Legacy PXRD pattern matching¶

POST /xrd/poscar_to_xyz — Convert POSCAR to XYZ¶