Jev directory
CommunityJev Models & Reimplementations
edceleste-decider 0.8B GGUF
mradermacher's static GGUF quantizations of IqSantiagow's 0.8B edceleste decider for llama.cpp, from Q2_K (0.5 GB) to f16 (1.6 GB), with Q4_K_M and Q4_K_S marked fast and recommended; no imatrix quants are planned by this quanter.
- Reading time
- 2 min
- Published by
- Community
- Author
- mradermacher
- Last verified
- Oct 2026
- Capabilities
- Open sourceSelf-hostedRuns locally
Highlights
- Twelve static GGUF files from Q2_K (0.5 GB) to f16 (1.6 GB) of IqSantiagow/edceleste-decider-0.8b.
- Q4_K_S and Q4_K_M are flagged fast, recommended; Q6_K very good quality; Q8_0 fast, best quality; Q3_K_M is flagged lower quality.
- No imatrix or weighted quants are planned by this quanter, per the card's note about its model requests page.
- Apache-2.0; this quantization card publishes no evaluation results of its own.
Watch out
Card reviewed 2026-10-17; this is mradermacher's generic static-quant card: it lists files, sizes and quality notes but publishes no evaluations, so quality claims for a 0.8B decider in low-bit files are unsupported and likely lossy. The upstream edceleste-decider card was not fetched for this digest.
Questions about this entry
Short answers drawn from the linked project or page.
A set of static GGUF quantizations by mradermacher of IqSantiagow/edceleste-decider-0.8b, a 0.8B decision model tagged for Elite Dangerous, packaged for llama.cpp under Apache-2.0.
Q2_K 0.5 GB, Q3_K_S 0.5, Q3_K_M 0.6, Q3_K_L 0.6, Q4_K_S 0.6, IQ4_XS 0.6, Q4_K_M 0.6, Q5_K_S 0.7, Q5_K_M 0.7, Q6_K 0.7, Q8_0 0.9 and f16 1.6 GB, sorted by size rather than quality.
The card marks Q4_K_S and Q4_K_M as fast and recommended, Q6_K as very good quality, Q8_0 as fast with best quality, and Q3_K_M as lower quality. For a 0.8B model the smallest Q2 and Q3 files carry the most risk.
No. The card says weighted and imatrix quants seem not to be available from this quanter at this time and that it has probably not planned them; requests go through the model-request discussion page.
No. There are no benchmark numbers, latency figures or accuracy tables on this card; it is a generic quantization release page with file links, quality notes and a thanks section.
IqSantiagow/edceleste-decider-0.8b, tagged decision-model, system-one and elite-dangerous. This digest fetched only the quant card, so upstream training and data details are outside this entry's grounding.
The upstream model is tagged system-one and decision-model, so it belongs to the Jev-adjacent decision-model family, but this card documents no serving contract or API; runtime behaviour would follow the upstream repository.
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
X: Jev picks which Apple HIG guides go into the agent's context. NatashaTheRobot built a small AskHIG CLI and…
X: Three decision models, one game of Snake
More Jev resources like this
prithivMLmods' GGUF builds of autotrust/JEV-35B, the Qwen3.5-35B-A3B MoE decision model that returns calibrated choice, noul and score answers over text and images; files run from BF16 to Q3_K_L for llama.cpp.
Jev Models & Reimplementations#community#open-models#local
Community
A calibrated image-capable 27B decision model that fits an 8 GB GPU: the JEV LoRA-and-head recipe trained on Prism ML's ternary Bonsai 2, returning typed probabilities with no generated text.
Jev Models & Reimplementations#community#open-models#local
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
A 4-bit AWQ plus GPTQ build of AutoTrust's JEV-35B from AI Outfitters: INT4 weights with FP8 and INT8 activations, calibrated on all 256 experts, and served by vLLM on a single 24 GB GPU.
Jev Models & Reimplementations#community#open-models#local
Community
A 4-bit W4A8 AWQ plus GPTQ build of AutoTrust's JEV-27B-VL from AI Outfitters that keeps the rank-32 LoRA decision head and a bf16 vision tower, served by vLLM on a single 24 GB GPU.
Jev Models & Reimplementations#community#open-models#local
Community
Mapika's NVFP4 fine-tune of Gemma-4-31B-it answers typed Choice, Noul and Score questions with calibrated probabilities and no generated text, self-reporting a 64.27 chance-corrected public Decision Index and 39 ms median latency on a B300.
Jev Models & Reimplementations#community#open-models#local
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.