1× DGX Spark
53tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
Local AI · Recipe · 1× DGX Spark
Bonsai-27B Q4_K_M with llama.cpp on 1× DGX Spark
by vcruz305
Intelligence (original model) · Artificial Analysisno independent value
≤ 3 bit: quantization may cost quality
Engine
- Engine
- llama.cpp 62061f91 (prism) / mainline cecbf5fb0
- Quantization
- GGUF Q4_K_M (imatrix); Ladder Q3_K_M–Q8_0
- Model family
- Bonsai-27B
- Context
- 16,384
- Parameters
- 27B dense
- Creator
- vcruz305
- GitHub stars
- 1
- Repo updated
- Jul 14, 2026
Measurements
Every sourced value of this recipe, each with its condition and source. Bars relative to the largest value in the group.
No sourced measurements for this recipe.
What you need
- Hardware
- 1 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightsvcruz305/Bonsai-27B-GGUF Apache-2.0
- Draft modelprism-ml/Bonsai-27B-gguf Apache-2.0
- Baseprism-ml/Bonsai-27B-gguf Apache-2.0
- Engine
- llama.cpp 62061f91 (prism) / mainline cecbf5fb0
- Context
- 16,384 tokens
Notes
What matters before you rebuild it.
- ≤ 3 bit: quantization may cost qualityAt 3 bit and below the model may answer noticeably worse than the original. The intelligence number refers to the original.
Sources
Related recipes
1× DGX Spark
48.7tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligence40with thinking
- Access on request
- License unclear
1× DGX Spark
40.9tok/sEveryday
1 request, prose prompt, prompt 50 tokens, with DSpark (per recipe) Source
Intelligence25with thinking
- License unclear
1× DGX Spark
32tok/sEveryday
1 request, prose prompt, with Drafting aus dem Prompt (Edit-Setup, lt. README 'drafts its guesses from your prompt') (per recipe) Source
Intelligence40with thinking
- License unclear
1× DGX Spark
30.2tok/sEveryday
1 request, story prompt, prompt 18 tokens, with MTP (per recipe) Source
Intelligence34with thinking
- ≤ 3 bit: quantization may cost quality
1× DGX Spark
26.4tok/sEveryday
1 request, prose prompt, with MTP (NEXTN) (per recipe) Source
Intelligence40with thinking
- Custom kernel required
- License unclear
Local AI in your company?
In a workshop we work out which models and which hardware fit your tasks, and build the first agent on your infrastructure.