aibuildai v2.0.0

aibuildai turns a dataset and a task description into a trained model.
v2.0.0 adds account sign-in; aibuildai run now requires a Pro plan.

Changes

• Sign in with aibuildai login (opens a browser).
• aibuildai run requires a Pro plan — see https://www.aibuildai.io/#products.
• Also new: whoami, logout, account.

Requirements

• Linux x86_64. Tested on Ubuntu 18.04+ and CentOS 8+; built against glibc 2.31.
• NVIDIA GPU with CUDA 12.x runtime.
• 16 GB RAM. ~5 GB disk for the binary.
• conda on PATH — aibuildai creates a Python env per task.
• Pro plan for aibuildai run.
• Anthropic API key in ANTHROPIC_API_KEY.

Getting started

curl -L -o aibuildai.tar.gz \
  https://github.com/aibuildai/AI-Build-AI/releases/download/v2.0.0/aibuildai-linux-x86_64-v2.0.0.tar.gz
tar -xzf aibuildai.tar.gz
cd aibuildai-linux-x86_64-v2.0.0
./install.sh

aibuildai login
aibuildai whoami

export ANTHROPIC_API_KEY=sk-ant-...
aibuildai run \
  --task-name house-prices \
  --data-dir /path/to/data \
  --instruction "Predict SalePrice from the training rows. Optimize RMSE." \
  --playground-dir /path/to/playground

aibuildai v0.1.1

Highlights

• Broad Linux compatibility: now supports systems with glibc 2.28+, including Ubuntu 18.04+, CentOS 8+, Debian 10+, and most GPU cloud instances.
• Binary size reduced from 206 MB to 108 MB.
• Bug fixes.

System Requirements

• Linux x86_64 (glibc >= 2.28)
• Anthropic API key

aibuildai v0.1.0

Initial public release of the autonomous ML engineering framework.

What's included

• Linux x86_64 binary package
• Multi-agent ML pipeline: setup, design, code, tune, revise, aggregate
• Interactive TUI and non-interactive CLI modes

Installation

curl -L -O https://github.com/aibuildai/AI-Build-AI/releases/latest/download/aibuildai-linux-x86_64-v0.1.0.tar.gz
tar -xzf aibuildai-linux-x86_64-v0.1.0.tar.gz
cd aibuildai-linux-x86_64-v0.1.0
./install.sh

Requirements

• Linux x86_64
• Anthropic API key (export ANTHROPIC_API_KEY=your-key)
• A prepared dataset directory and task instruction

See the README for full usage examples.