qcinput

qcinput reads an xyz or gjf geometry, applies task-specific keywords from a TOML config, It reads an xyz or gjf geometry, applies task-specific keywords from a TOML config, and emits ORCA or Gaussian input files for int, ts, or sp jobs.

Install With pip

pip install qcinput

For development (from source):

pip install -e .

Usage

qcinput generate [<path/to/structure.xyz|.gjf> ...] [--input-dir structures/] [-c|--config <path/to/qcinput.toml>] [-o output.inp | --output-dir outputs/]

Compatibility shorthand (same behavior):

qcinput [<path/to/structure.xyz|.gjf> ...] [--input-dir structures/] [-c|--config <path/to/qcinput.toml>] [-o output.inp | --output-dir outputs/]

Show version:

qcinput --version
qcinput -V

First-time setup:

qcinput init-config
qcinput init-config --kind int
qcinput init-config --kind ts
qcinput init-config --kind sp

init-config always writes int/ts/sp task sections in one TOML file; --kind only sets the initial active value of [qcinput].kind.

Default config path:

./qcinput.toml

You can override this with:

QCINPUT_CONFIG=/path/to/config.toml qcinput water.xyz

Example:

qcinput water.xyz -c qcinput.toml -o water_opt.inp

Batch examples:

qcinput *.xyz *.gjf -c qcinput.toml
qcinput --input-dir structures/ -c qcinput.toml
qcinput *.xyz --input-dir structures/ -c qcinput.toml --output-dir generated/

*.xyz / *.gjf are expanded by your shell before qcinput runs. --input-dir scans the given directory for .xyz and .gjf files. --output-dir writes all generated files into one directory and creates it automatically. -o/--output is only valid when converting a single input structure.

For ORCA and Gaussian, kind = "ts" now uses a two-step TS template. Use:

qcinput water.xyz -c qcinput.toml -o water_ts.inp

ORCA TS settings:

[orca.task.ts]
step1_keywords = ["Opt"]
step2_keywords = ["OptTS", "Freq"]
constraint_atoms = [[0, 1], [2, 3]]
calc_hess = true

constraint_atoms accepts either one pair ([0, 1]) or multiple pairs ([[0, 1], [2, 3], ...]). ORCA TS uses global [orca].nprocs and [orca].maxcore for the compound workflow. Step-2 reads <output_stem>_Compound_1.xyz.

Gaussian TS settings (Link1):

[gaussian.task.ts]
step1_keywords = ["Opt=ModRedundant"]
constraint_atoms = [[0, 1], [2, 3]]
step2_keywords = ["Opt=(TS,CalcFC,NoEigenTest,NoFreeze)", "Freq", "Geom=AllCheck", "Guess=Read"]

constraint_atoms accepts one pair ([0, 1]) or multiple pairs ([[0, 1], [2, 3], ...]) and will be converted to B i j F lines. Note: TOML uses 0-indexed atoms, but Gaussian ModRedundant is 1-indexed; conversion is automatic. Legacy modredundant is still supported for compatibility. %chk is auto-generated as <structure_stem>.chk (e.g. water.xyz -> water.chk). This applies to Gaussian int, sp, and ts.

When qcinput.engine = "gaussian", default output suffix is .gjf.

Config Format (TOML)

We use TOML because it is readable for humans, easy to version control, and strongly structured.

[qcinput]
engine = "orca" # or "gaussian"
kind = "int" # int | ts | sp

[molecule]
charge = 0
multiplicity = 1

[orca]
nprocs = 8
maxcore = 4000
extra_keywords = ["TightSCF", "NormalSCF"] # optional, appended at end
smd = false # optional
smd_solvent = "toluene"

[orca.task.int]
base_keywords = ["B3LYP", "def2-TZVP"]
keywords = ["Opt", "Freq"]

[gaussian]
nprocshared = 8
mem = "8GB"
extra_keywords = ["SCF=Tight"] # optional, appended at end of keywords

[gaussian.task.int]
base_keywords = ["B3LYP/def2TZVP"]
keywords = ["Opt", "Freq"]

Output Snippet

# Generated by qcinput v<version> (https://github.com/yushengyangchem/qcinput)
! Opt Freq B3LYP def2-TZVP
%pal
  nprocs 8
end
%maxcore 4000
* xyz 0 1
O 0.000000 0.000000 0.000000
H 0.757000 0.586000 0.000000
H -0.757000 0.586000 0.000000
*