User repository on GitHub
Run System One decision models from Hugging Face locally behind the Jev / System One API, Ollama style
- Forks
- 5
- Last pushed
- Updated Oct 2026
Repo stats from the GitHub API, cached Oct 2026.
Jev directory
CommunityJev Tools & CLIs
Ollajev
An Ollama-style local runner for System One decision models from Hugging Face: pull, serve and query them behind the same /v1/systemone API as TypeSafe's Jev, with a terminal model manager and a browser playground.
- Stars
- 16
- Language
- Python
- License
- APACHE-2.0
- Reading time
- 3 min
- Published by
- Community
- Author
- nvkudva
- Last verified
- Oct 2026
- Languages
- Python
- Capabilities
- Open sourceRuns locally
Highlights
- Drop-in for typesafe-sdk clients: point TYPESAFE_BASE_URL at the local server and the routes, requests and response shapes stay the same.
- Version 0.4 makes Vega, a trained physics engine on a frozen Qwen3.5, the default in 0.8B (1.8 GB) and 4B (9.5 GB) sizes with a 73,728-token context.
- Ollama-style commands — pull, list, run, serve, ps, rm — plus a terminal model manager and a playground that compares answers across models.
- Images, audio and video can ride beside the state with Vega, Clef-Omni, d1-omni and Clef on Apple Silicon through MLX.
- The installer compiles llama.cpp locally — Xcode Command Line Tools on macOS, a C/C++ compiler on Linux — and models stay in the Hugging Face cache.
Quickstart
bash
curl -fsSL https://raw.githubusercontent.com/nvkudva/ollajev/main/install.sh | sh
ollajev serveWatch out
Grounded in the truncated repository README and GitHub metadata reviewed on 2026-10-15; the README's example request block is cut off mid-JSON in what was reviewed, and the 20-40x CPU speedup claim comes from the vendor's 0.4 release notes rather than an independent benchmark.
Questions about this entry
Short answers drawn from the linked project or page.
A local server for System One decision models, in the style of Ollama: it runs models pulled from Hugging Face on your machine behind the same /v1/systemone API as TypeSafe's Jev, with no generated text — just calibrated probabilities per typed question.
On macOS or Linux, curl -fsSL https://raw.githubusercontent.com/nvkudva/ollajev/main/install.sh | sh, or brew install nvkudva/tap/ollajev, or uv tool install ollajev. Windows uses a PowerShell one-liner. On first run ollajev opens a model manager.
It is a drop-in for typesafe-sdk: point TYPESAFE_BASE_URL at the local server and the same routes and request and response shapes apply. The playground at 127.0.0.1:8000/playground lets you compose a state and typed questions in the browser.
Vega, a trained physics engine on a frozen Qwen3.5, in 0.8B (1.8 GB) and 4B (9.5 GB, :4b) sizes with a 73,728-token context and image input. Install with brew install nvkudva/tap/ollajev, then ollajev pull and ollajev serve.
Yes, beside the state: Vega, Clef-Omni, d1-omni and Clef support media input on Apple Silicon through MLX. The playground can attach an image, audio clip or video when building a request.
The installer compiles llama.cpp: macOS needs the Xcode Command Line Tools and Linux a C/C++ compiler. Config and logs stay in ~/.ollajev, while models stay in the Hugging Face cache — run ollajev rm before uninstalling to free them.
No key is required; it runs locally. Idle models make room for a new one, and Qwen3.5-based PyTorch models are documented as answering 20-40x faster on CPU, though that figure comes from the project's release notes rather than an independent benchmark.
More Jev resources like this
A correctness-validated Laya runtime for Apple Silicon that runs upstream Laya on the MLX GPU and the Apple Neural Engine, using the ANE only after an artifact passes a parity gate on your machine, plus a local Jev-compatible server.
Jev Repos & SDKs#community#catalog-import#madewithjev
Community
tc3oliver/laya-applePython
tc3oliver/laya-applePythonToolkit, knowledge base and local dashboard for building with Jev via OpenRouter Alpha Decisions or the TypeSafe Direct API: a zero-dependency SQLite request log, a localhost REST server and Python CLI demos.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
rileycarney/jevtoolsHTML
rileycarney/jevtoolsHTMLA lightweight Python engine that mimics Jev's structured decisions on local LLMs via Ollama, vLLM or llama.cpp: choice, score and noul questions come back as JSON with probabilities and confidence, plus an OpenAI-compatible gateway.
Jev Repos & SDKs#community#catalog-import#logicrw
Community
tapsin/jev-localPython
tapsin/jev-localPythonTurns an inbox into a short action queue: Jev answers seven typed questions per email thread in one request, plain Python rules decide what enters the queue, and cached answers mean changing a rule never needs new API calls.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
tusharck/jev-inbox-queuePython
tusharck/jev-inbox-queuePythonOllaya's site for its local decision-model runtime: one binary pulls open checkpoints and serves them on TypeSafe-compatible endpoints, now at Ollaya 0.8.0 with an app and command line for macOS, Windows, Linux, and Docker.
Jev Tools & CLIs#community#local#runtime
Community
A desktop app for dictation, audio-file transcription and meeting notes on Windows, macOS and Linux, using local Whisper models or the OpenAI API, with an opt-in MCP server and local history API so agents can search saved transcripts.
Jev Tools & CLIs#community#catalog-import#logicrw
Community
knuckles92/openwhisperPython
knuckles92/openwhisperPythonFrom the Jev guides
Original write-ups on what this entry is for.
From the Jev AI model community
Posts from builders shipping with Jev right now.