AI Models
Reference data

What's actually running on your phone.

"On-device AI" in Ovels isn't one model — it's a small roster, chosen automatically based on your phone's hardware, with a cloud fallback behind all of it.

The tier ladder
TierModel familyRole
MiniQwen 3 · 0.6BFastest replies, lowest memory footprint — the default on older or storage-constrained devices
MidLlama 3.2 · 1BA middle ground between the mini and full coaching models
FullQwen 3 · 1.7BThe most capable on-device coach, for newer hardware
VisionLFM 2.5-VL · 450MNames a dish from a meal photo for offline photo-mode recognition
One model, three hardware backends

Each tier isn't a single file — it's compiled for whichever accelerator your phone actually has, chosen automatically:

ExecuTorch

The general-purpose backend, works on any supported device as the baseline.

Core ML

Routes inference through the iPhone's Neural Engine for faster, more efficient replies.

QNN

Routes inference through the phone's dedicated NPU on supported Qualcomm Android hardware.

Bring your own model

Advanced users can also upload a custom compatible model file rather than use one of the bundled tiers — treated as its own tier alongside Mini, Mid, and Full.

If no on-device model has finished downloading yet, the app doesn't block — it runs on a basic, rules-only mode until a tier is ready, then upgrades automatically. See AI Coach for how the active tier is used.

← Back
Nutrition Database
Next →
FAQ