# kairos-agi/kairos

Official code for world model Kairos

Repository: https://github.com/kairos-agi/kairos
Canonical: https://ross.abutalabs.com/products/kairos
Language: Python
License: Apache-2.0
License Family: permissive
Last push: 2026-07-08T03:29:15+00:00

## Health v2 (maintenance only)
Score: 57/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 91, release rhythm 35, longevity 19
- inputs: {"age_days": 274, "days_push": 56, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2598, forks 504 (observed 2026-08-28T04:07:03.185813+00:00)

## What it is
Kairos is the official open-source implementation of a 4B-parameter native cross-embodiment world model that unifies video understanding, future visual generation, and executable robot action prediction. It provides inference code and pretrained weights for robot manipulation benchmarks and supports real-time edge-side deployment for embodied AI.

## Use cases
- predict future video states for robot manipulation
- generate executable robot actions from world model
- run world model inference on robot benchmarks like RoboTwin and LIBERO
- deploy embodied AI models on edge devices
- experiment with cross-embodiment pretraining for robotics
- evaluate long-horizon reasoning in embodied agents

## When to choose
- you need a unified world model that predicts both future video and robot actions
- you want pretrained cross-embodiment weights for manipulation benchmarks
- you need efficient edge-side deployment for embodied AI

## When to avoid
- you need a general-purpose video generation model without robot action prediction
- you lack GPU resources for 4B-parameter model inference
- you need a mature production robotics stack rather than research code

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-inference, video-processing, simulation
- domain: artificial-intelligence, robotics, machine-learning, computer-vision
- platform: python
- tags: world-model, cross-embodiment, robot-learning, action-prediction, video-generation, embodied-ai, linux, gpu

## Member repositories
- kairos-agi/kairos (main) score 57

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:03.185813+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:21:20.817652+00:00, confidence not recorded.
  - readme: https://github.com/kairos-agi/kairos (fetched 2026-08-28T04:07:03.185813+00:00, sha be27c7665323)
- Data as of 2026-08-30T08:39:29.467469+00:00.