Standard Kernel
Standard Kernel Co. is an AI infrastructure company building a system that automatically generates optimized low-level kernel code — the programs that execute computations on GPUs and TPUs. It aims to eliminate the kernel-engineering bottleneck across the AI stack by combining LLMs with program analysis to produce hardware kernels (matmul, attention, fused feed-forward networks) that match or exceed hand-tuned libraries such as cuBLAS and FlashAttention3, including reported results at 99–116% of cuBLAS on H100 BF16 matmul. Backed by a $20M seed round with participation from General Catalyst, the company is early-stage and does not yet publish a public API, SDK, or developer platform.
Standard Kernel is profiled on the APIs.io network. Tagged areas include Company, AI Infrastructure, GPU, Kernel Optimization, and Compilers.
Standard Kernel’s developer surface includes engineering blog and 3 more developer resources.
Kin Score
Security Posture 1
Authentication, domain security, vulnerability disclosure, and trust-center signals.
Resources
Access & Security 1
Authentication, authorization, and security posture
Company 2
The organization behind the API
Other 1
Properties that don't map to a standard resource type