Compare projects

Select 2–4 projects to compare on the same basis.

Show differences only

KernelBench: Can LLMs Write GPU Kernels? - Benchmark + Toolkit with Torch -> CUDA (+ more DSLs)

A streamlined and customizable framework for efficient large model (LLM, VLM, AIGC) evaluation and performance benchmarking.

Popularity

Repository stars1,3183,658Highest

Fork196554Highest

Maintenance

Last push2026-03-242026-11-06

Maintenance notesNot verifiedNot verified

Repository created2024-10-252023-12-07

License & use

LicenseLicense not declaredApache-2.0 · Permissive

PricingPricing not verifiedPricing not verified

Basics

Browse by purposeModel and data evaluationModel and data evaluation · Retrieval-augmented generation

TypeFrameworks & SDKsFrameworks & SDKs

SubcategoryGeneralGeneral

PlatformNot verifiedNot verified

First collected2026-11-032026-11-03

LanguageJupyter NotebookPython

License tiers are derived from metadata. Check obligations for your use; unknown does not mean prohibited. Measured at 2026-11-06 UTC