Jev directory
CommunityJev Models & Reimplementations
Kev-9B Artemis GGUF
TurinTech's independent GGUF conversion of Jared Palmer's Kev-9B decision model; the default IQ3_M file is 4.78 GB and matches the BF16 original on 92.2% of 2,000 test decisions with a minus 1.2-point accuracy change.
- Reading time
- 2 min
- Published by
- Community
- Author
- TurinTech
- Last verified
- Oct 2026
Used inJev for Software & DevTools
- Languages
- Shell
- Capabilities
- Open sourceSelf-hostedRuns locally
Highlights
- Independent TurinTech GGUF conversion of jaredpalmer/kev-9b, a 9B decision model; default IQ3_M is 4.78 GB.
- IQ3_M matches BF16 on 92.2% of 2,000 test decisions (minus 1.2 points) with no confident-answer changes.
- IQ4_XS (4.87 GB, minus 1.1), Q3_K_L (4.45 GB, minus 0.7) and Q3_K_S (4.07 GB, minus 2.2) cover tighter memory budgets.
- Measured 9.1 GiB peak GPU memory for the default file; tested on Apple M3 Max Metal and an NVIDIA RTX 5090.
- Public ggml-org Q4_K_M (6.36 GB) is closer at 93.2% same answers but is 1.5 to 2.3 GB larger than the Artemis files.
Quickstart
bash
llama-server -hf turintech/Kev-9B-Artemis-GGUF:IQ3_M --port 8080 -c 16384 -ub 4096 -b 4096Watch out
Card reviewed 2026-10-17; the 92.2% same-answer figure and accuracy deltas are TurinTech's own measurements on a 2,000-decision selection panel and, per the card, do not establish performance on unseen inputs. The conversion is independent and unendorsed by the Kev authors, and measured probabilities shift slightly with batch size.
Questions about this entry
Short answers drawn from the linked project or page.
An independent TurinTech conversion of jaredpalmer/kev-9b into GGUF files for llama.cpp; the default is IQ3_M at 4.78 GB, a 73% size cut from the 17.9 GB BF16 original.
Kev accepts text with a set of typed questions and returns a probability for every allowed answer in a single pass instead of generating text, served through llama-server's /v1/systemone endpoint.
IQ3_M 4.78 GB (92.2% same answers, minus 1.2 points, default), IQ4_XS 4.87 GB (91.6%, minus 1.1), Q3_K_L 4.45 GB (88.8%, minus 0.7) and Q3_K_S 4.07 GB (87.1%, minus 2.2); no confident answers changed in any file.
ggml-org Q8_0 (9.5 GB) matches the original 98.7% of the time at minus 0.2 accuracy and Q4_K_M (6.36 GB) 93.2% at minus 0.6; TurinTech's files are 1.5 to 2.3 GB smaller and use 1.4 to 2.0 GiB less GPU memory.
Use llama-server -m Kev-9B-Artemis-IQ3_M.gguf --port 8080 -c 16384 -ub 4096 -b 4096 with llama.cpp from 9 October 2026 or newer; the card measured 9.1 GiB peak GPU memory for the default file.
They cover 2,000 test decisions on the card's selection panel and, per the card, do not establish performance on unseen inputs; the conversion is also independent and not endorsed by the Kev authors.
A 4,914-token input was answered at both batch settings in the card's testing, and an optional -np 8 -c 32768 --cache-ram 0 configuration serves eight slots with 4,096 tokens each at about 0.6 GiB more GPU memory.
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
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
A community Q8_0 GGUF of Mapika/decider-0.8b, the smallest System One typed-decision model: one forward pass returns calibrated Choice, Score and noul answers, reporting 0.405 on JevBench hard-public at 920 MB resident.
Jev Models & Reimplementations#community#open-models#local
Community
Fluid Inference's fixed-shape Core ML conversion of the classification path of Fastino's GLiNER2.5-Decide for Apple silicon: the DeBERTa-v3-large encoder and shared label classifier score up to four decision heads per call, in fp16, 8-bit or 6-bit packages.
Jev Models & Reimplementations#community#open-models#local
Community
An 8-bit MLX conversion by SirSahOl of chaoliangUNSW/Jev-Style-0.8B-Decision-v3 built for Apple Silicon native GPU inference: about 990 MB of active VRAM, a 262,144-token context and an average of 8.25 bits per weight.
Jev Models & Reimplementations#community#open-models#local
Community
A Q8_0 GGUF of knowlet's Gemma-4-12B-Unified-System-One with a native System One decision runtime and an independently fitted GGUF temperature, shipped with an F16 mmproj for the model's native image and audio projections.
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.