Base Optimization Stack: From Open Weights to Frontier On-Device Inference Speed
• 4
AI & ML interests
local AI, GPU kernels, performance optimisation, agentic research pipelines, automated research
Recent Activity
Articles
Organization Card
Base Compute is an AI inference lab.
Our mission is to run AGI on-device.
We believe a world where people have access to intelligence that runs on hardware they control is fundamentally better.
A small senior team in Melbourne and Berlin working at the runtime and silicon level, where the hard problems are.
We're hiring -> https://www.basecompute.co/careers
models 41
basecompute/gpt-oss-120b
Text Generation • Updated
basecompute/NVIDIA-Nemotron-3-Nano-30B-A3B
Text Generation • Updated
basecompute/GLM-5.2
Text Generation • Updated
basecompute/gpt-oss-20b
Text Generation • Updated
basecompute/Qwen3.8-27B
Text Generation • Updated • 274
basecompute/Muse-Glimmer-30B
Image-Text-to-Text • Updated • 22 • 1
basecompute/Qwen3-0.6B
Text Generation • Updated • 24
basecompute/whisper-large-v3-turbo
Automatic Speech Recognition • Updated • 12
basecompute/whisper-large-v3
Automatic Speech Recognition • Updated • 15
basecompute/whisper-large-v2
Automatic Speech Recognition • Updated • 11
datasets 0
None public yet