# nesaorg/nesa

Run AI models end-to-end encrypted.

Repository: https://github.com/nesaorg/nesa
Canonical: https://ross.abutalabs.com/products/nesa
Homepage: https://nesa.ai
Language: Python
License Family: other
Topics: ai, encryption, privacy, deep-learning, llms
Last push: 2025-02-10T03:16:49+00:00

## Health v2 (maintenance only)
Score: 24/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 6, release rhythm 35, longevity 44
- inputs: {"age_days": 619, "days_push": 569, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3167, forks 246 (observed 2026-08-28T04:07:46.718914+00:00)

## What it is
Nesa is a privacy-preserving AI inference platform that runs models like Llama, Mistral, and Stable Diffusion with end-to-end encryption using its Equivariant Encryption technique, so the serving party never sees input or output data. It offers a ChatGPT-compatible API and is backed by a decentralized Layer-1 blockchain network for verifiable, distributed AI execution.

## Use cases
- run llm inference on sensitive data without exposing it
- private encrypted ai inference via api
- replace on-prem ai infrastructure with a privacy-preserving api
- verify ai inference results on-chain
- run stable diffusion or llama models encrypted
- decentralized ai compute network

## When to choose
- you need strong privacy guarantees for AI queries over sensitive data
- you want a ChatGPT-compatible API with encrypted inference and minimal code changes
- you want verifiable, decentralized AI execution instead of trusting a centralized provider

## When to avoid
- you need a fully self-hosted, air-gapped solution with no external network dependency
- you require an OSI-approved open-source license for commercial use
- you need guaranteed low-latency inference and cannot tolerate any encryption overhead

## Facets
- artifact type: service
- maturity: active
- function: llm-inference, machine-learning, cryptography, security, privacy, api-framework
- domain: artificial-intelligence, large-language-models, privacy, blockchain, machine-learning, apis
- platform: python, cloud, self-hosted
- tags: encrypted-inference, equivariant-encryption, decentralized-ai, chatgpt-compatible-api, layer-1-blockchain, private-ai, web-server

## Member repositories
- nesaorg/nesa (main) score 24

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:07:46.718914+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 2026-08-30T07:25:09.491315+00:00, confidence not recorded.
  - readme: https://github.com/nesaorg/nesa (fetched 2026-08-28T04:07:46.718914+00:00, sha 84a35cc984de)
  - homepage: https://nesa.ai (fetched 2026-08-29T09:39:54.681515+00:00, sha e745c03eaf87)
  - site_page: https://docs.nesa.ai/nesa (fetched 2026-08-29T09:39:54.684095+00:00, sha e6044b9a795a)
  - site_page: https://nesa.ai/about (fetched 2026-08-29T09:39:54.686004+00:00, sha bc5b3734689a)
  - site_page: https://nesa.ai/pricing (fetched 2026-08-29T09:39:54.687618+00:00, sha 56e556028580)
- Data as of 2026-08-30T08:39:29.467469+00:00.