Skip to content
repository radar
Models & Inference

☁️skypilot

ACTIVE STEADY

skypilot-org/skypilot · homepage ↗

The AI Compute Platform for frontier teams. SkyPilot turns fragmented AI compute into one AI supercomputer, so frontier AI teams build custom intelligence faster.

⭐ Very popular: 10k stars, gaining about 41 a week

View on GitHub ↗

repo profile

vintage 2021 5 years old
language Python
license Apache-2.0

momentum

total stars 10k
stars added last week +41
used by 353 repos & packages
commits / week 31 cooling · -31% vs prior mo
issues closed 90% ~13d to close, median
contributors 276 (+5 in 15d)
release cadence fortnightly
last activity today

durability

backing foundation-backed UC Berkeley Sky Computing Lab
openness permissive
bus factor 3 concentrated
top-author share 37% 6 mo

bus factor = how many people it takes to cover more than half the commits (6 months). 1 is a solo project; higher means the work is spread across a team. top-author share is the single busiest author's slice of those commits.

since we covered it

monthly average + 182/mo (+2%/mo) · + 1k total since PR#25

why it's a big deal

  • Teams running AI workloads get one interface across Kubernetes, Slurm, and 20+ clouds, so jobs stay portable instead of tied to a single provider.
  • It targets cost and reliability directly with automatic failover across providers, cost-aware scheduling, and autostop for idle resources.
  • The BYOC model keeps all compute inside the user's own accounts and clusters, which matters for teams with existing infra and data-residency constraints.

under the hood

  • Workloads are declared in a YAML or Python spec covering resource requirements, setup commands, and tasks; SkyPilot then handles provisioning, file syncing, dependency install, and execution.
  • The codebase is roughly 89 percent Python with JavaScript frontend components, exposing both a CLI and an API server for shared team access.
  • It layers scheduling features on top of existing schedulers, including gang scheduling and multi-cluster support on Kubernetes and multi-node jobs with binpacking.

Radar summary, generated from the project's public sources

star history

PR#25 · 9k10k now Aug 2021Aug 2026
  1. PR#25 9k 2026-01-14
  2. now 10k + 1k since first covered

curve is sampled from GitHub's star history, plus our own daily readings since we covered it; the dashed stretch is before we first covered it, the solid line since. figures at coverage are the numbers we printed then (approx.), current count is live.

understory

Deep in the understory, shipping hard, still flying under the radar.

+38 understory score output 91 · clout 53
Aug 2025 Jul 2026
  • output, commits & releases
  • clout, star velocity

output = commits & releases; clout = star velocity, both 0 to 100 monthly indices; the gap where output runs above clout is the understory. The understory →

covered in

  • PR#25 2026-01-14 below the radar

    Unified AI infrastructure orchestration

similar projects

compare these →
  • 🧩 TensorZero

    Rust

    TensorZero is an open-source LLMOps platform that unifies an LLM gateway, observability, evaluation, optimization, and experimentation.

    12k OFF THE RADAR
  • 👁️ supervision

    4.7× the stars

    We write your reusable computer vision tools. 💜

    49k ACTIVE
  • ⌨ MLflow

    2.6× the stars

    The open source AI engineering platform for agents, LLMs, and ML models. MLflow enables teams of all sizes to debug, evaluate, monitor, and optimize production-quality AI applications while controlling costs and managing access to models and data.

    28k ACTIVE

comments

Sign in with GitHub to add your blip on skypilot.

loading comments…