generated: '2026-08-04'
method: searched
source: >-
  npm, PyPI, RubyGems, crates.io, pkg.go.dev registry probes; github.com/Essential-AI;
  huggingface.co/EssentialAI; ollama.com/library/rnj-1; lmstudio.ai
note: >-
  Essential AI publishes NO client SDK for any language — there is no first-party HTTP API to write
  one against. Registry probes for essential-ai / essentialai / @essentialai/* on npm and PyPI all
  returned 404. What the company does ship is a Homebrew cask for its RNJ macOS desktop app and
  open-weight model distributions. No `SDKs` pointer is wired in apis.yml because no client library
  exists; wiring one would fabricate an ergonomics signal the provider has not earned.
packages:
- language: shell
  registry: homebrew
  name: rnj
  tap: Essential-AI/homebrew-essentialai
  url: https://github.com/Essential-AI/homebrew-essentialai
  cask: https://raw.githubusercontent.com/Essential-AI/homebrew-essentialai/main/Casks/rnj.rb
  install: brew install --cask Essential-AI/essentialai/rnj
  version: 1.0.0
  description: RNJ — instant-access AI chat that runs privately on your Mac (macOS >= Ventura, depends on ollama)
  artifact: https://storage.googleapis.com/homebrew-essentialai/RNJ-1.0.0.zip
  official: true
  kind: desktop-app
- language: n/a
  registry: ollama
  name: rnj-1
  url: https://ollama.com/library/rnj-1
  install: ollama run rnj-1
  tags: [latest, 8b]
  description: Rnj-1 8B open-weight model in the Ollama library, trained from scratch by Essential AI (requires Ollama >= 0.13.3)
  official: false
  official_note: hosted in the curated Ollama library namespace; trained and released by Essential AI but not published from an Essential AI account
  kind: model-distribution
- language: n/a
  registry: lm-studio
  name: essentialai/rnj-1
  url: https://lmstudio.ai/models/essentialai/rnj-1
  description: Rnj-1 in the LM Studio model catalog
  official: false
  official_note: third-party catalog mirroring the Hugging Face release
  kind: model-distribution
registries_probed:
- {registry: npm, queries: [essential-ai, essentialai, "@essentialai/sdk"], found: 0}
- {registry: pypi, queries: [essential-ai, essentialai, eai-taxonomy], found: 0}
- {registry: homebrew, queries: [Essential-AI/homebrew-essentialai], found: 1}
models:
  source: https://huggingface.co/EssentialAI
  note: first-party open-weight releases; the canonical Essential AI distribution channel
  items:
  - {id: EssentialAI/rnj-1, license: apache-2.0, pipeline: text-generation}
  - {id: EssentialAI/rnj-1-instruct, license: apache-2.0, pipeline: text-generation}
  - {id: EssentialAI/rnj-1.5-instruct, pipeline: text-generation}
  - {id: EssentialAI/rnj-1-instruct-GGUF, note: quantized GGUF build for llama.cpp/Ollama/LM Studio}
  - {id: EssentialAI/eai-distill-0.5b}
datasets:
  source: https://huggingface.co/datasets?other=EssentialAI
  items:
  - {id: EssentialAI/essential-web-v1.0, note: document-level web dataset organized by a twelve-category taxonomy}
  - {id: EssentialAI/eai-taxonomy-code-w-dclm}
  - {id: EssentialAI/eai-taxonomy-stem-w-dclm}
  - {id: EssentialAI/eai-taxonomy-med-w-dclm}
  - {id: EssentialAI/eai-taxonomy-math-w-fm}
  - {id: EssentialAI/rnj-1-base-evals}
  - {id: EssentialAI/rnj-1-instruct-evals}
source_code:
  org: https://github.com/Essential-AI
  first_party_repos:
  - {name: eai-taxonomy, url: 'https://github.com/Essential-AI/eai-taxonomy', language: Python}
  - {name: reflection, url: 'https://github.com/Essential-AI/reflection', language: Python, license: MIT}
  - {name: maxtext-external, url: 'https://github.com/Essential-AI/maxtext-external', language: Python, license: Apache-2.0}
  - {name: rnj-1-assets, url: 'https://github.com/Essential-AI/rnj-1-assets'}
  - {name: homebrew-essentialai, url: 'https://github.com/Essential-AI/homebrew-essentialai', language: Ruby}
x-evidence:
  fetched: '2026-08-04'
  github_api: https://api.github.com/orgs/Essential-AI/repos
  huggingface_api: https://huggingface.co/api/models?author=EssentialAI