How can we help?

Browse our docs, join the community, or reach out directly.

Search docs, tutorials, and FAQ...

Frequently asked questions

Is Pexper really free?
Yes. Pexper is free during the alpha. Bring your own API keys or use local models at no cost. We don't take a cut of your AI usage — you pay providers directly at their standard rates.
Do I need API keys to use Pexper?
No. You can connect your existing ChatGPT Plus or Claude Pro subscription directly, or run local models with Ollama, LM Studio, or Apple MLX — no API key required.
What platforms does Pexper support?
Pexper is a desktop app for macOS, Windows, and Linux. Apple Silicon (M-series) is recommended for running local models efficiently.
Where are my API keys stored?
All API keys are stored locally on your device only. They're never transmitted to Pexper's servers. When you run a workflow, your prompts go directly from your device to the provider you choose.
Can I run Pexper completely offline?
Yes. Use local models via Ollama, LM Studio, Apple MLX, or llama.cpp, and Pexper will work 100% offline. No data leaves your machine.
What happens to the workflows I build?
Everything stays local. Your workflows, prompts, and generated content are saved on your device. You own all of it.
How do I report a bug or request a feature?
Open an issue on our GitHub repo, post in Discord, or email support@pexper.com. During alpha, we respond within 24 hours.
Will Pexper stay free?
The alpha is free forever. We may introduce a paid version in the future — if we do, it will be a one-time purchase, not a subscription. Alpha users will be notified with at least 30 days of notice.

Oxel

Pexper's companion app for on-device image processing — background removal, upscaling, and optimization.

Common Oxel questions

How do I download the ML models?
Open Oxel (or Pexper), go to Settings → Models, and tap Download on the model you need. Models are fetched from pexper.com on demand. No manual installation required.
What's the difference between Real-ESRGAN x2 and x4?
x2plus upscales to 2x resolution, x4plus to 4x. Both are ~33 MB and run on Apple Silicon via Core ML. Use x2 for faster processing or subtle enhancement, x4 for maximum upscaling of small images.
Does Oxel run offline?
Yes. Once the models are downloaded, everything runs on-device using Apple Core ML. Your images never leave your Mac. An internet connection is only needed for the initial model download.
What are the system requirements?
macOS 14 Sonoma or later. Apple Silicon (M1 or later) is strongly recommended — Core ML models run significantly faster on the Neural Engine than on Intel Macs.
Can I use Oxel's models in Pexper?
Yes. OxelKit is the shared framework between both apps, so upscale and background removal models downloaded in one will work in the other. In Pexper, they're exposed as Action Cards for on-device image processing.

Still need help?

We typically respond within 24 hours during alpha.