Jev directory
CommunityJev Tools & CLIs
Flide
Flide is an open-source inference engine for Jev-style decision models that serves the Jev /v1/systemone API. Exact state forking makes it up to 9.5x faster than vLLM and SGLang. It runs on NVIDIA GPUs under Linux or Windows WSL2 and ships precompiled kernels for RTX 40 and 50 series cards.
- Reading time
- 2 min
- Published by
- Community
- Added
- Nov 2026
Highlights
- Serves the Jev /v1/systemone API for decision-model inference.
- Exact state forking makes it up to 9.5x faster than vLLM and SGLang.
- Runs on NVIDIA GPUs under Linux or Windows WSL2; installed from PyPI with uv.
- Precompiled kernels for RTX 40 series, RTX 50 series, and RTX PRO 6000 GPUs.
- Email triage demo tags 10,000 emails with eight questions each in 28 seconds on one H200.
Questions about this entry
Short answers drawn from the linked project or page.
Flide is published on PyPI. Create a Python 3.12 environment with uv venv -p 3.12, activate it, then run uv pip install flide torch. It runs on an NVIDIA GPU under Linux or Windows WSL2, and the docs cover every model, images, the API, and all options.
Flide runs on an NVIDIA GPU under Linux or Windows WSL2. Precompiled kernels ship for RTX 40 series, RTX 50 series, and RTX PRO 6000 GPUs. With --quantize fp8, Gemma 4 12B fits on a single 24 GB card.
Flide serves the Jev /v1/systemone API. As of version 0.1.9, /v1/systemone accepts several requests in one body.
Exact state forking is Flide's inference technique: every answer is the one the model gives the question alone, and it is up to 9.5 times faster than vLLM and SGLang. The flide verify command checks exact forking on your own model and requests.
Flide 0.1.0 served Qwen3.5 and Gemma 4 chat models plus the OneJev and Open-Jev decision models. Later releases added many decision model families, including GLiDE, KnowLine, Sieve, Drex, Jeffan, JevAny, and d1-3B from LiquidAI.
Flide is Apache-2.0 licensed and maintained by the OmniJev organization on GitHub, where the repository has 477 stars, 31 forks, and 63 commits.
Since Flide 0.1.2, a question can take up to 256 options.
Reactions & coverage
Posts, threads, and videos about this entry from around the web.
X: Standard One SH (StandardThinking)
LinkedIn: What multimodal decision models actually change, and what they do not.
LinkedInJevDirectory.org
Hacker News: Decision Index 0.2.1 scores open reproductions of Jev on a frozen suite.
jackbravo4 pointsDecision Index 0.2.1 – 70 open reproductions of TypeSafe Jev's Decision ModelRead the thread on Hacker NewsHacker News: Pointer to a DiffusionGemma plus vLLM tutorial for a Jev-shaped classifier.
1 pointHow to build a Jev-style classifier with DiffusionGemma and vLLMRead the thread on Hacker NewsMore Jev resources like this
A proof-of-concept HTTP service for constrained decisions with a local MLX model: send text or JSON plus typed choice, score, or noul questions and it returns probabilities over the answers you declared — no generated text, no parsed model JSON.
Jev Tools & CLIs#community#github#python
Community
snellingio/system-onePython
snellingio/system-onePythonA local web app where Claude Code runs the hunt and Jev judges every competitor post, surfacing the ones most likely to work for your audience. It adds saved-post analysis, transcription, an Ask AI view and a My Studio planning page.
Jev Tools & CLIs#community#github#python
A local, non-autoregressive security guardrail suite powered by Convai Innovations' Laya model, returning typed decision probabilities for security inspection. It positions itself against Jev on latency and against generative LLMs on hallucination risk.
Jev Tools & CLIs#community#github#security
Windows desktop overlay for Feishu/Lark group chats that syncs messages through lark-cli, uses Jev to score four MBTI dimensions, and draws scroll-tracking name badges over the client. MIT licensed, Python 3.11+.
Jev Tools & CLIs#github#python#local
Communityakiiwa-games/lark-mbti-jev
A local workbench for System One-style models on Apple Silicon Macs: it serves Laya through MLX or Core ML behind a Jev-compatible /v1/systemone endpoint, proxies other Jev-compatible servers like Jeff and Kev, recommends models and evaluates them on Japanese data.
Jev Tools & CLIs#community#github#python
Community
katya4oyu/systemone-workbenchPython
katya4oyu/systemone-workbenchPythonA self-hosted server that removes ads from podcasts before you hit play: it transcribes locally with faster-whisper or Parakeet, detects and cuts ad segments with your choice of LLM or System One, and learns sponsors across episodes.
Jev Tools & CLIs#community#github#python
Community
ttlequals0/minuspodPython
ttlequals0/minuspodPythonFrom 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.