Using Orca Slicer in CLI Mode for Automated G-code Generation

[LukaCek]

Hi everyone,

I'm working on an application that allows users to upload an STL file, and the program will automatically generate a .gcode or .gcode.3mf file using Orca Slicer. My goal is to run Orca Slicer in command-line interface (CLI) mode, similar to how PrusaSlicer and SuperSlicer can be used with CLI commands.

I've been searching for documentation or examples on how to use Orca Slicer in CLI mode, but I haven't found much information so far. Does anyone know if Orca Slicer supports CLI commands for slicing, and if so, what the proper syntax would be?

I am using a Bambu P1 printer, and this setup is intended for it. Additionally, if anyone knows of any other programs that would allow me to achieve this programmatically, I would appreciate the recommendations.

If you've had experience using Orca Slicer in an automated workflow, I'd love to hear your insights on how to set it up.

Thanks in advance!

[Ale-Ponte]

Hi @LukaCek ,
I came across your post while researching the same topic (using OrcaSlicer via command line to automate slicing tasks) and I noticed you didn’t receive a response.

I'm currently working on a similar goal: building an automated system to generate slicing previews (print time, material usage, etc.) from .stl files uploaded via a web interface. Ideally, the slicer would run headlessly with predefined profiles and export a .gcode or .3mf or even just slicing data.

I’ve tried various combinations of CLI commands (based on what’s available from BambuStudio and Orca forks), but so far, I haven't had success executing a full slicing pipeline via terminal on Windows 10, at least.

Did you ever make any progress on this since your original post?
Or has anyone else figured out a viable way to run OrcaSlicer headless or through scripts?

Would be great to reopen the discussion if anyone is still interested.

Thanks!

[AFKFelix]

Hey @Ale-Ponte, @LukaCek,
I'm currently developing a small API to slice models using Orca Slicer over a REST API.
If you want, you can check it out here, it is still in early development, so feedback is very welcome.

To answer your question for the CLI itself, it is not too complicated once you get the basics.
The -h option generally provides a good overview of the commands.
You call the Orca Slicer binary with the model or .3mf project file as last argument, and you can use --slice with 0 for all plates or the plate number to slice a specific plate as G-code.
Passing both the --slice and the --export-3mf <file-name> will produce a .gcode.3mf file containing the G-code, which is used by Bambu Lab printers.
You can also load profiles as JSON files and do things like auto-orient or arrange. The options for this are well described in the help command.
Using the --curr-bed-type option and the name of the bed type, you can specify the type of bed, e.g. Textured PEI Plate.

I hope this helps and feel free to ask other/more specific questions if u need :)

[plop-dev]

@AFKFelix I've just started using your API too, (great work btw) and I do think a JSON response for /slice would be best, with a url to download the gcode. However, add a download url would mean an extra endpoint, which means you need to track slice data, and add ids, etc. I'm currently keep res.download() with header 'X-Slice-Metadata' containing a base64 string of all the slice data I want to export:

const metaHeader = Buffer.from(JSON.stringify({ times, filament})).toString('base64');
res.setHeader('X-Slice-Metadata', metaHeader);

res.download(gcodes[0]);

[AFKFelix]

@plop-dev Yes thanks, good idea with the header. I think I'll add something similar to this next.

[pterob]

a JSON response would be ideal since it's the most used format when it comes to returning REST responses @AFKFelix

[AFKFelix]

@pterob

a JSON response would be ideal since it's the most used format when it comes to returning REST responses @AFKFelix

You are correct, the problem with this is that the slice route is originally meant to return the sliced gcode/3mf. If I were to return a JSON response including the metadata now I would need to include a download URL for the file, which would mean that I would need to store the sliced result and also manage its deletion and so on. This would make the endpoint more complex and break the existing API. That's why for now I will stick with the header approach, since it is lightweight, simple and does not affect the file response. I may add this functionality later, but for now I will stick with the simpler approach of the headers.

[pterob]

fair enough, let's stick to whatever works that's convenient for you

[LukaCek]

Hey — I made a small web tool that automatically slices uploaded STL files using OrcaSlicer (repo: https://github.com/LukaCek/slice_api
, demo: https://printprice.cekluka.com/
). I wanted to share how I’m driving OrcaSlicer from the command line and what outputs it produces so others can reproduce or correct my approach.

How I invoke OrcaSlicer (conceptually)

I call the OrcaSlicer binary with a single command that:

loads a printer/profile and filament definitions,

optionally auto-arranges / auto-orients the model on the plate,

slices the model (plate 0 in my case),

exports slice metadata and a .gcode.3mf (Bambu-friendly) file.

A compact example of the kind of command I run (adjust paths and quoting per OS):

orcaslicer
--arrange 1
--orient 1
--export-slicedata ./temp
--load-settings "printers/Bambu Lab P1S 0.4 nozzle.json;process/ @bbl X1C.json"
--load-filaments test_data/filament.json
--slice 0
--export-3mf ./temp/output.3mf
--info
temp/file.stl

Quick notes on the flags:

--load-settings (printer + process) is how I inject the selected quality/profile.

--load-filaments points to filament definitions so the slicer can report usage.

--export-slicedata tells OrcaSlicer to write out slice metadata next to the 3mf.

--slice 0 slices the whole plate.

--export-3mf produces the .gcode.3mf that Bambu printers consume.

What OrcaSlicer produces (what I read back)

When the run completes, I expect (and parse) these outputs:

result.json — contains return_code and error_string for quick failure detection (e.g. model too big/small or other errors).

output.3mf (the file I request with --export-3mf) — this is a zip-like archive that can be renamed to .zip and extracted.

Inside the extracted 3mf you typically see:

Metadata/slice_info.config (XML) — contains filament info: type, used_m, used_g (useful to compute material cost).

Metadata/plate_1.gcode — contains slicer comments with timing estimates and other fields, e.g.:

model printing time: 1h 16m 56s;

total estimated time: ...

first layer printing time = ...

max_z_height: ...
These comment lines are what I regex-parse to get model/total/first-layer times and max Z for pricing and reporting.

Practical needs — what I’m asking for

What I really want to understand is how to import and control slicing settings from the CLI:

How do people import filament definitions reliably? Is --load-filaments path/to/file.json the recommended approach, or do you normally embed filament data in a profile/3mf? If you use JSON files, what structure do you use and can you share a tiny example?

How can I override or change process settings (temperatures, fan, support generation, infill %, layer height, etc.) from the command line? Is there a standard way to:

pass a different process/profile JSON, or

apply one-off overrides (e.g., set nozzle temp = 210, infill = 20%) without editing profile files manually?

If the usual approach is to load a printer + process profile, what is the exact format for --load-settings when combining printer + process (especially when profile names contain spaces or @bbl style suffixes)? Examples showing correct quoting/escaping would be very helpful.

Do people typically:

keep a library of JSON profiles and --load-settings them, or

programmatically generate a temporary .3mf/profile and pass that to OrcaSlicer?

If anyone can share minimal, working examples (Linux/Windows) showing:

importing a filament file,

loading a process/profile and changing a couple of parameters from CLI,

Thanks — Luka (author of slice_api)

[paprago]

@LukaCek I am trying build the same automated slicing pipeline but when i try to run the commands on the command prompt nothing happens. No error is shown too. I want to do it locally and i am using windows. Is this a windows problem?

[AFKFelix]

@paprago This might actually be a Windows issue from what I know. Try the latest version, I heard it works with that. Otherwise you can run the Linux version using WSL or using a Docker container.

[paprago]

@AFKFelix Thank you for the information! I will try with the latest version else shift to the Linux version.

[paprago]

@LukaCek @AFKFelix i have got everytihing running now. But the slicing works only when i use the json file shared by @LukaCek on Git hub. When i try yto replace it with my custom presets it fails saying preset is invalid. Need some help on the way to generate correct config json for printer, process and filament.

[AFKFelix]

@paprago From my experience the profile stuff is a bit tricky. The best way I found is to put it together manually based on the profile parts and inheritance from the profiles of the resources folder of OrcaSlicer.

I am looking for a better way of doing this so if anyone has any suggestions, I would love to hear them.