OpenJev 27B — VRAM & /v1/systemone setup
Written by Jakub Rusinowski · Last updated
A 27B decision model for text, web pages and screenshots, up to 52 options a question. llama.cpp's server runs it natively at /v1/systemone from the ggml-org GGUF. CC-BY-NC-4.0: non-commercial only.
OpenJev 27B needs about 20 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache. On Apple Silicon that figure comes out of unified memory.
Call OpenJev 27B
Served by llama-server (llama.cpp) (default port 9931). This model does not run in Ollama.
- Question types
- choice noul (yes / no) score
- Inputs
- text image
- Max candidates per question
- 52
- Serve with
llama-server (llama.cpp)- Endpoint
/v1/systemone
llama-server -hf ggml-org/OpenJev-GGUFhttp://localhost:9931/v1/systemone
Setup for servers other than Ollama →
Build a request for this model Decision models guide
Related guides: Servers other than Ollama · Confidence, thresholds and calibration
Other decision models: Nimble 9B · Tev1 4B · Tev1 0.8B · Winnow 12B · Winnow E4B · Decider 2B · Decider 4B · Decider 35B-A3B (NVFP4) · Decider 12B Vision · Decider 31B v2 (NVFP4) · JevK5 4B · Intern-Decision 4B · AutoJev 27B · Laya · Clef 27B · Clef-Flash 9B · Rune 26B-A4B v3 · OpenDecider-nano · OpenDecider-small · OpenDecider-small MLX 4-bit · Bosun v3.1 0.6B · Bosun v3.1 1.7B · d1-3B · d1-omni-600M · THX-01 322M · mammoth-1 27B · elephant-1 9B · Strands Decider 2B (hobson-v21) · kev 4B · kev 9B
Hardware fit
Weights plus overhead plus the KV cache at a 8,192-token prompt, on NVIDIA RTX 4090 (24 GB). A publisher build is sized from its file; the other rows are modelled at a standard quant. Decision requests are short, so no long-context figure is shown. Image tokens are not modelled: each image adds to the prompt, so a request with images needs more memory than these rows.
| Quant | Memory | VRAM | Fit |
|---|---|---|---|
| BF16 Publisher build · 53.81 GB file | 56.4 GB | Too big | |
| Q4_K_M 4.83 bpw · modelled quant | 21.9 GB | Tight | |
| Q6_K 6.56 bpw · modelled quant | 25 GB | Offload | |
| Q8_0 8.50 bpw · modelled quant | 31.1 GB | Offload |
Published file size: Q4_K_M 19.4 GB · Q8_0 28.6 GB · BF16 53.81 GB. A download size from the model publisher — not a VRAM requirement.
Not the ZefanCai Open-Jev project: llama.cpp's openjev family is converted from openjev/openjev. Images need the multimodal projector (mmproj).
Specifications
Verified — Checked against the primary source — the model card or the vendor spec page — and corroborated by a second independent source.
- Parameters
- 27.36B
- Context window
- Not published
- Architecture
- Fine-tune of openjev base (27B)
- Provider
- openjev (community)
- Licence
- cc-by-nc-4.0
- Availability
- Weights downloadable
- Practical deployment
- Single consumer GPU
- Specified at
- Q4_K_M
- System RAM
- 42 GB
- Record updated
- 2026-10-25
- Runs with
- llama.cpp: supported source
Research / non-commercial only — this licence does NOT permit shipping a commercial product.
OpenJev 27B — frequently asked questions
What is OpenJev 27B?
OpenJev 27B is a decision model: you send it a state and typed questions (choice, yes/no/unknown, or a score) and it returns one answer per question with a probability for every option. It is not a chat model.
How do I run OpenJev 27B locally?
Served by llama-server (llama.cpp) (default port 9931). This model does not run in Ollama. Start it with: `llama-server -hf ggml-org/OpenJev-GGUF`.
How much memory does OpenJev 27B need?
About 20 GB for the weights plus overhead at Q4_K_M, before the prompt's KV cache. Decision prompts are short, so the cache stays small.