Jebadiah 27B
Jebadiah 27B loads in 16.8 GB at Q4_K_M, comfortably from 32 GB of memory, Macs included, starting at the MacBook Pro M5 Pro 48GB. Decision checks are validated up to 4,096 tokens. Below is the memory math we ran for it, the smallest card that fits, and the fastest machine we track.
- Parameters
- 27B
- Format
- Q4_K_M
- Min memory
- 32 GB
- Best for
- Decisions
How much memory does Jebadiah 27B need?
Jebadiah 27B is a 27B dense model built for decisions. The weights alone take 16.8 GB at Q4_K_M; add context and the OS, and it belongs in the 32 GB memory class. Every number below comes from the ModelFit engine: speeds are estimates, labeled est., because we do not benchmark hardware ourselves.
Run with llama.cpp: llama-server -hf frontier-infra/jebadiah-27b-v2-1-GGUF:Q4_K_M -c 4096. Ollama and LM Studio are not verified for this build, so this page does not give commands for them.
KV = fp16 estimate (q8_0 cache roughly halves it). "Comfortable" = weights + KV within the engine's tiered budget (~70-85% of memory). Context figure is the validated 4,096-token window.
Which machines run Jebadiah 27B?
- Smallest GPU
- AMD Radeon RX 7900 XTest. 37 tok/s · 20 GB VRAMtypical error 50%
- Fastest
- NVIDIA GeForce RTX 5090est. 89 tok/stypical error 5%
- Runs on a Mac?
- MacBook Pro M5 Pro 48GBest. 17 tok/stypical error 10%
Verdicts come from the same engine that powers our fit checker: memory capacity first, bandwidth for speed. How we compute these numbers · What typical error means · Hugging Face model card
Go deeper
- Also tracked: Jebadiah 27B (Q5)(Q5_K_M, 19.5 GB, min 32 GB)
- Also tracked: Jebadiah 27B (Q8)(Q8_0, 29 GB, min 48 GB)
More Jebadiah models
1 pagesGet told when a better model fits your MacBook Air M5 16 GB
Get one email when a new open-weight model is a better fit for this machine.
For: MacBook Air M5 16 GB
You'll also get the weekly ModelFit email. By signing up you agree to our Privacy Policy. Unsubscribe from either email anytime, in one click.
Frequently asked questions
How much memory does Jebadiah 27B need?
16.8 GB for the Q4_K_M weights, plus ~0.3 GB of KV-cache at 4k context, about 17.1 GB total. Comfortable from 32 GB of VRAM or unified memory. That 4,096-token window is the validated decision context, not the base model's longer config window.
Does Jebadiah 27B run on a Mac?
Yes, from the MacBook Pro M5 Pro 48GB (est. 17 tok/s, typical error 10%). Unified memory means the RAM budget is the only limit.
How do I run Jebadiah 27B?
With llama.cpp: llama-server -hf frontier-infra/jebadiah-27b-v2-1-GGUF:Q4_K_M -c 4096. Use the decision script in the GGUF repo rather than a chat prompt. Ollama and LM Studio are not verified for this build.
What is the smallest GPU for Jebadiah 27B?
The AMD Radeon RX 7900 XT (20 GB VRAM) is the smallest tracked card that runs Jebadiah 27B comfortably, est. 37 tok/s, typical error 50%.
Cite this page
ModelFit: Jebadiah 27B, specs, memory math and hardware verdicts. https://modelfit.io/models/jebadiah-27b/ (dataset updated 2026-10-30, CC BY 4.0).