Razer’s open-source AI development toolkit adds image, video and audio model support, expands Arm64 compatibility, and underpinned the AVA Mini campaign with Akash Network at lower per-image inference costs.
Image courtesy of Razer
IRVINE, CALIF. –
Razer said on April 30 that it has released an updated version of Razer AIKit, its free, open-source AI development toolkit. The update expands AIKit beyond text-based workflows to include support for image, video and audio AI models, while adding broader hardware compatibility, including Arm64 architectures.
The toolkit was first introduced at CES 2026. According to the company, the latest version is designed to allow developers to build, run and deploy AI models on local hardware rather than relying solely on cloud-based services. AIKit manages GPU discovery, configuration and performance tuning, giving developers greater control over deployment environments and costs.
The update adds support for image-generation models, including Tongyi-MAI/Z-Image-Turbo and FLUX.2-klein-base-4B. Razer said the toolkit can run across supported local and edge systems, including recent Razer Blade laptops, as well as selected Arm64-based professional AI platforms.
Razer also said AIKit was used to power AVA Mini, a 3D AI pet companion campaign launched in March 2026 as part of its AVA desk companion project. The campaign allowed users to upload pet photos and receive AI-generated character images.
For the campaign, Razer partnered with Akash Network, a decentralized GPU marketplace, to deploy AIKit across distributed RTX 4090 and RTX 5090 GPUs through AkashML. The campaign ran from March 31 to April 4 and generated more than 11,000 images, with an average turnaround time of 3.24 seconds per image and peak throughput of 30 images per minute.
Razer said the deployment reduced inference costs to US$0.01 per image, compared with conventional cloud API costs that it estimated at US$0.03 to US$0.15 per image. The company and Akash Network also published a joint white paper outlining the campaign’s system architecture and unit economics.
AIKit is now available through Akash Console, allowing users to deploy and manage AIKit workloads through a browser-based interface. Razer said future releases are expected to add voice and video support.
Dan Yoo
댓글
0