Google AI Edge
Run AI on-device with Google AI Edge: MediaPipe, LiteRT-LM and LiteRT across platforms.
At a glance
Price not confirmed · Source review 2026-11-07- Price & billing
No usable pricing facts were read for this stack in this pass; the developer tools (MediaPipe, LiteRT, LiteRT-LM) are published as free developer resources with no paid tier observed on the page read.
- Free allowance
No usable pricing facts were read for this stack in this pass; the developer tools (MediaPipe, LiteRT, LiteRT-LM) are published as free developer resources with no paid tier observed on the page read.
- Best suited to
Developers who want to run AI on-device (mobile, web, desktop, embedded) using Google's MediaPipe, LiteRT and LiteRT-LM stack.
Fit is editorial guidance, not a measured result.- Platforms
Access: not separately verified. Operating systems not separately verified.
Published-source facts and editorial guidance. No product test is implied. Confirm current conditions before purchase.
No product screenshot is claimed for this listing. Use the official website ↗ to inspect the current interface.
What to know before choosing
Good fit
Developers who want to run AI on-device (mobile, web, desktop, embedded) using Google's MediaPipe, LiteRT and LiteRT-LM stack.
Limits and trade-offs
It is a developer toolkit, not a consumer product; no pricing facts were obtained for it in this pass.
Plans and free allowances
No usable pricing facts were read for this stack in this pass; the developer tools (MediaPipe, LiteRT, LiteRT-LM) are published as free developer resources with no paid tier observed on the page read.
Evidence
Official homepage ↗ (reviewed 2026-11-07), Official pricing page ↗ (reviewed 2026-11-07)
Official published documentation reviewed; not a product test
Dates: Last reviewed 2026-11-07. Last reviewed is the latest source or editorial review date; it is not a content-change date.
A closer look at Google AI Edge
Google AI Edge is Google's end-to-end stack for building and deploying on-device ML and AI across platforms. It includes MediaPipe task APIs for prebuilt features (background blur, face landmarks, object detection), LiteRT-LM to run LLMs on-device across Android, iOS, web and embedded devices, and LiteRT hardware-accelerated runtimes for custom models converted from PyTorch, JAX, TensorFlow and Keras. The same technology powers live translation on Pixel, dynamic backgrounds on Google Meet and creator effects on YouTube Shorts. The AI Edge Portal benchmarks on real Android devices, and Model Explorer visualises model architectures.
Editorial guidance based on published documentation. The workflow below is a suggested starting point, not a measured product result.
When to choose Google AI Edge
- A developer wants to add prebuilt AI features such as background blur or object detection to an app.
- A developer wants to run an LLM on-device across Android, iOS and web.
- A developer wants to deploy a custom model with hardware acceleration.
When to choose something else
- A non-developer looking for a ready-made AI app.
- A user who only needs cloud AI with no on-device requirement.
A practical starting workflow
- Pick the starting point: a MediaPipe task API or a custom model.
- Integrate the task API, or convert the custom model for LiteRT deployment.
- Run the model on-device across Android, iOS, web or embedded targets.
- Benchmark with the AI Edge Portal and debug with Model Explorer.
Check before committing
- It is a developer platform (SDKs, task APIs, runtimes), not an end-user app.
- The /pricing URL redirects to a Google AI Studio page and carries no usable pricing facts for this stack in this pass.
Common questions
What is in the Google AI Edge stack?
MediaPipe (prebuilt task APIs), LiteRT-LM (run LLMs on-device) and LiteRT (hardware-accelerated runtimes for custom models).
Which platforms does it run on?
Android, iOS, web, desktop and embedded/IoT, across CPU, GPU and NPU.
How much does it cost?
Unknown in this pass: the /pricing URL fetched a Google AI Studio page with no usable pricing facts for this stack.
Compare approaches for this task
Sources and checks
Last source review is not a date of product testing. No rating or test result is implied by inclusion. Check the current plan and terms before buying.