wowsuch ai
Live

GoogleDeveloper platform

ML Drift

ML Drift is an open-source, cross-platform GPU compute engine for on-device AI/ML inference, available as a standalone library and as the core GPU acceleration engine within LiteRT. It supports GPU backends including OpenGL ES, OpenCL, Metal, and WebGPU, with desktop previews for Windows and Linux.

What it does

Developers can use it to build and accelerate on-device AI workloads, including real-time vision, generative AI, and custom GPU compute graphs.

Availability
Open-source release; standalone C++ APIs and LiteRT packages available
Pricing model
Apache 2.0 license
Deployment
On-device across Android, iOS, web, and desktop
Privacy
Not established
Licence
Apache 2.0
Capability
analysis
Modality
textimageaudiovideomultimodal
Access
SDKopen weights
Deployment
on-device
Pricing
open-sourcefree
Target user
developerresearcherenterprise

Evidence

6 claims
  • ML Drift is available as a standalone library and as the GPU acceleration engine within LiteRT. [1]
    Serving as the core GPU acceleration engine within LiteRT , ML Drift is also available as a standalone library for custom graphics and inference runtimes providing a unified foundation to deliver peak performance everywhere.
    Google. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • ML Drift supports OpenGL ES, OpenCL, Metal, and WebGPU. [1]
    By abstracting hardware and low-level API complexities of on-device GPUs across OpenGL ES, OpenCL, Metal, and WebGPU, ML Drift empowers developers to build real-time, interactive ML experiences from advanced video effects to generative AI across multiple platforms.
    Google. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • ML Drift supports 5D tensors out of the box. [1]
  • ML Drift can run on Android, iOS, web, Windows, Linux, and macOS.
  • ML Drift dynamically switches kernels and layout configurations between LLM prefill and decode stages. [1]
    ML Drift dynamically switches kernels and layout configurations depending on the active execution phase.
    Google. Quoted text is outside CC BY 4.0 and keeps its original rights.
  • The legacy TFLite GPU delegate will no longer receive new feature updates. [1]
    With the launch of ML Drift, the legacy TFLite GPU delegate will no longer receive new feature updates.
    Google. Quoted text is outside CC BY 4.0 and keeps its original rights.

Sources

1
1
GoogleOwner-controlled
https://developers.googleblog.com/ml-drift-next-gen-gpu-aiml-inference-at-the-edge/ Retrieved 2026-10-29 00:21 UTC
Verified

Versions

9
VersionEffectiveChangeChanged fieldsState
92026-10-29 00:21 UTCmaterial changecapability, pricing, availability, pricing modelCurrent
82026-10-27 00:19 UTCmaterial changemodality, access, pricing, availability, deploymentCurrent
72026-10-21 10:01 UTCmaterial changecapability, modality, access, pricing, availability, deploymentCurrent
62026-10-20 10:59 UTCmaterial changecapability, modality, access, availability, deploymentCurrent
52026-10-16 14:32 UTCmaterial changecapability, modality, pricing, availability, pricing model, deploymentCurrent
42026-10-14 16:15 UTCmaterial changecapability, access, pricing, availability, pricing model, deploymentCurrent
32026-10-13 17:05 UTCmaterial changecapability, access, pricing, availabilityCurrent
22026-10-09 20:45 UTCmaterial changecapability, access, pricing, availability, pricing model, deploymentCurrent
12026-10-08 21:36 UTCpublicationNoneCurrent

In briefs

2

Open in the catalog

Original catalog content is licensed under CC BY 4.0, attributed to wowsuchai. Source quotations and linked content keep their original rights.