4× RTX PRO 6000
81.7tok/sEveryday
1 request, prose prompt, with MTP Source
Intelligence34with thinkingno thinking 22
- ≤ 3 bit: quantization may cost quality
Local AI · Recipe · 2× DGX Spark
GLM-5.2 753B NVFP4 with vLLM on 2× DGX Spark
by tonyd2wild
Intelligence (original model) · Artificial Analysisno independent value
≤ 3 bit: quantization may cost quality
Engine
- Engine
- vLLM 0.24.0
- Quantization
- 2-bit MoE planes (pruned 208) + NVFP4 attention
- Model family
- GLM-5.2
- Context
- 32,768
- Parameters
- 753B MoE
- Creator
- tonyd2wild
- GitHub stars
- 39
- Repo updated
- Sep 11, 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
- 2 × NVIDIA DGX Spark (GB10)
- Weights
- Main weightssapidlabs/GLM-5.2-2bit-MoE-planes-pruned208-tp2 MIT
- Draft modelRedHatAI/GLM-5.2-speculator.dspark Apache-2.0
- Basezai-org/GLM-5.2-FP8 MIT
- Engine
- vLLM 0.24.0
- Context
- 32,768 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.
- Experimental
Sources
Related recipes
4× RTX PRO 6000
70.7tok/sEveryday
1 request, prose prompt, with MTP Source
Intelligence34with thinkingno thinking 22
- ≤ 3 bit: quantization may cost quality
4× RTX PRO 6000
64tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligence34with thinkingno thinking 22
- ≤ 3 bit: quantization may cost quality
- Custom kernel required
4× RTX PRO 6000
58.5tok/sEveryday
1 request, prose prompt, with MTP (per recipe) Source
Intelligenceno independent value
- Custom kernel required
4× RTX PRO 6000
104tok/sMixed
1 request, mixed prompt set, with MTP (per recipe) Source
Intelligence34with thinkingno thinking 22
- ≤ 3 bit: quantization may cost quality
4× RTX PRO 6000
92.5tok/sMixed
1 request, mixed prompt set, with MTP Source
Intelligenceno independent value
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.