Jev directory
CommunityJev Models & Reimplementations
Ara-Jev-Faisal GGUF
mradermacher's static GGUF quants of Omartificial-Intelligence-Space/ara-jev-Faisal, an Arabic intent, policy and routing classifier; Q4_K_S and Q4_K_M are marked fast and recommended, under a non-commercial CC-BY-NC-4.0 licence.
- Reading time
- 2 min
- Published by
- Community
- Author
- mradermacher
- Last verified
- Oct 2026
- Capabilities
- Self-hostedRuns locally
Highlights
- Static GGUF quants of Omartificial-Intelligence-Space/ara-jev-Faisal, from Q2_K at 2.1 GB up to f16 at 8.8 GB, plus 0.5-0.8 GB mmproj supplements.
- Upstream tags describe an Arabic decision model for intent classification, routing and policy, with Gulf Arabic and code-switching, built on Qwen3.5 with LoRA.
- Q4_K_S and Q4_K_M are the recommended fast picks at about 2.7-2.9 GB; Q8_0 at 4.7 GB is fast with the best quality and Q6_K is called very good.
- Upstream metadata lists xnli, boolq_ar, MassiveIntentClassification, SANAD, labr, ASTD, OSACT4 and Arabic sarcasm and sentiment sets.
- Licensed CC-BY-NC-4.0, so commercial use is not covered; this quantization card itself publishes no benchmarks or usage example.
Watch out
Card reviewed 2026-10-17; this is a quantization card that documents no evaluations, so quality judgments rest only on the listed quant types and the upstream model's metadata, the licence is CC-BY-NC-4.0 which rules out commercial use, and the card provides no runnable command.
Questions about this entry
Short answers drawn from the linked project or page.
mradermacher's static GGUF quantization of Omartificial-Intelligence-Space/ara-jev-Faisal, an Arabic intent, routing and policy decision model. It ships as fourteen files covering Q2_K through f16 plus two multimodal projectors.
Q4_K_S at 2.7 GB and Q4_K_M at 2.9 GB are marked fast and recommended; Q6_K at 3.7 GB is very good quality and Q8_0 at 4.7 GB is fast with the best quality. f16 at 8.8 GB is called 16 bpw and overkill.
Two multi-modal supplements ship alongside the text quants: mmproj-Q8_0 at 0.5 GB and mmproj-f16 at 0.8 GB, for pairing vision input with the model in llama.cpp-based runtimes that support multimodality.
The upstream metadata lists Arabic and English. Its tags highlight Gulf Arabic and code-switching, which fits intent classification and routing on mixed-language support and policy traffic.
The base model metadata lists datasets including facebook/xnli, christykoh/boolq_ar, mteb/MassiveIntentClassification, arbml/SANAD, mohamedadaly/labr, Arabic sarcasm and sentiment sets, OSACT4 and Arabic reviews.
The upstream model is CC-BY-NC-4.0, so the quants inherit a non-commercial licence. The card states this in its metadata block, and there is no permissive or commercial licence on the repository.
No. The card says weighted and imatrix quants do not seem to be available at this time and suggests opening a community discussion to request them. The published files are static quants only.
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
The Q8_0 8-bit GGUF build of StartLux-Decision-v1-27B, 28.60 GB for llama.cpp, served through a decision wrapper that the card measures at 100% same decisions as the original and 209 of 231 on JevBench public.
Jev Models & Reimplementations#community#open-models#quantized
Community
The Q8_0 8-bit GGUF build of StartLux-Decision-v1-9B, 9.53 GB for llama.cpp, with the card reporting 100.0% same decisions as the original weights and 200 of 231 correct (86.6%) on JevBench public.
Jev Models & Reimplementations#community#open-models#quantized
Community
A small distilled sibling of Perplexity's pplx-decider-v1.2: a Qwen3.5-4B decider that reads a state plus typed choice, noul and score questions in one pass, keeping most of the 27B model's Decision Index.
Jev Models & Reimplementations#community#open-models#multimodal
Community
Perplexity's 27B decision model rebuilt on Qwen3.8-27B with a joint request format that answers a whole request's questions in one sequence, reporting 68.09 on Decision Index 0.3 and 77.43 on vision.
Jev Models & Reimplementations#community#open-models#multimodal
Community
TurinTech's independent GGUF conversion of the 27B Kev decision model for llama.cpp: the default IQ4_XS file is 13.91 GB and keeps 97.8% of the BF16 original's answers over 2,000 test decisions, with zero changed confident answers.
Jev Models & Reimplementations#community#open-models#local
Community
TurinTech's independent GGUF conversion of Jared Palmer's Kev-4B decision model; the default Q4_K_M file is 2.54 GB and returns the same answer as the BF16 original on 93.0% of 2,000 test decisions with no confident-answer changes.
Jev Models & Reimplementations#community#open-models#quantized
Community
From 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.