HeyGen x Google Cloud: Bringing Avatar IV to TPUs

Event type: PartnershipPrimary topic: Generative Image & Video Also: Cloud Infrastructure Also: ML Frameworks & Training

First reported OCT 3, 2026 · 00:49 UTC · 21d ago · 1 source

Crest summary — written by AI from the 1 source below; every figure is checked against them · model deepseek-chat

Summary

HeyGen ported its 18B+ parameter Avatar IV video generation model to Google Cloud's Trillium (v6e) TPUs using torchax and XLA.

Key facts

HeyGen moved Avatar IV, an 18B+ parameter video generation model, to Google Cloud TPUs.
The port used torchax and XLA on Trillium (v6e) TPUs.
The work was announced on the Google Developers Blog on 2026-10-03.

Why it matters

Running a large video generation model on TPUs gives builders an alternative hardware path to GPUs, and the torchax and XLA tooling shows how existing PyTorch models can be moved to that stack.

Companies

Technologies

XLAframework
Avatar IVmodel
Trilliumproduct
torchaxlibrary

Matched by rule, not by AI: a repository linked in the coverage, or a technology or company in this event that Crest tracks.

Repositories

openxla/xlaA machine learning compiler for GPUs, CPUs, and ML accelerators4.6K starsC++Matches a named technology

Original sources (1)

The report this event was built from. Crest of Tech keeps no article bodies: the full text is at each source.

Google Developers BlogOfficial · Primary
HeyGen x Google Cloud: Bringing Avatar IV to TPUsPublish time not given
Read at Google Developers Blog