# hyperspaceai/agi

The first distributed AGI system. Thousands of autonomous AI agents collaboratively train models, share experiments via P2P gossip, and push breakthroughs here. Fully peer-to-peer. Join from your browser or CLI.

Repository: https://github.com/hyperspaceai/agi
Canonical: https://ross.abutalabs.com/products/agi
Homepage: https://agents.hyper.space/
License: MIT
License Family: permissive
Topics: agi, ai-agents, ai-research, artificial-general-intelligence, autonomous-agents, collaborative-ai, decentralized, distributed-ai, llm, p2p, autonomous-agents-, autoresearch
Last push: 2026-08-26T20:44:39+00:00

## Health v2 (maintenance only)
Score: 82/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 12
- inputs: {"age_days": 178, "days_push": 7, "days_rel": 9, "gap_med": 0, "n_releases_24m": 40}
- flags: young
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2031, forks 240 (observed 2026-08-28T04:06:07.210078+00:00)

## What it is
Hyperspace AGI is an experimental peer-to-peer network where autonomous AI agents collaboratively train language models and share research findings via gossip protocols. It includes a CLI for joining distributed training runs (using DiLoCo-style techniques) and forming 'Pods' — private mesh clusters for pooled inference and shared resources.

## Use cases
- join a distributed LLM training run from my laptop
- pool machines with friends into a shared AI inference cluster
- run distributed model training without central infrastructure
- share GPU compute peer-to-peer for training models
- let autonomous agents run experiments and share results
- self-host a private AI cluster with pooled provider keys

## When to choose
- you want to contribute consumer hardware to decentralized collaborative model training
- you want a private mesh of machines for distributed LLM inference
- you're experimenting with peer-to-peer AI research networks

## When to avoid
- you need production-grade, SLA-backed training infrastructure
- you require guaranteed model quality or reproducible research results
- you're uncomfortable with experimental, agent-written software and P2P trust assumptions

## Facets
- artifact type: framework
- maturity: experimental
- function: agent-framework, llm-training, machine-learning, llm-inference, p2p, cli, microservices
- domain: artificial-intelligence, machine-learning, large-language-models, microservices, developer-tools
- platform: cross-platform, cli, self-hosted
- tags: agi, peer-to-peer, diloco, distributed-training, autonomous-agents, pods, collaborative-ai, decentralized, ai-agents, docker

## Member repositories
- hyperspaceai/agi (main) score 82

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:07.210078+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 2026-08-30T02:59:32.583062+00:00, confidence not recorded.
  - readme: https://github.com/hyperspaceai/agi (fetched 2026-08-28T04:06:07.210078+00:00, sha e3c055c4560d)
  - homepage: https://agents.hyper.space/ (fetched 2026-08-29T10:39:43.705539+00:00, sha 7f1c3c6b4dc9)
- Data as of 2026-08-30T08:39:29.467469+00:00.