← CanIRunLocal · All hardware · All models
Z.ai · 1.33B parameters · vision.
Ollama weight files: Q8_0 1.8 GB (glm-ocr:q8_0); BF16 2.7 GB (glm-ocr:latest).
Yes on 62 machines in this list. The smallest is M1 8GB: BF16, about 3.7 GB, roughly 20 tokens per second, via Ollama. ollama run glm-ocr:latest
ollama run glm-ocr:latestollama run glm-ocr:latestollama run glm-ocr:latestollama run glm-ocr:latestvllm serve zai-org/GLM-OCR --max-model-len 8192 --gpu-memory-utilization 0.9| Hardware | Runtime | Quant | Memory | Fit | Speed | Command |
|---|---|---|---|---|---|---|
| M1 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~20 tok/s | ollama run glm-ocr:latest |
| M2 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| M3 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| RTX 3060 Ti 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~100 tok/s | ollama run glm-ocr:latest |
| RTX 3070 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~100 tok/s | ollama run glm-ocr:latest |
| RTX 4060 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~60 tok/s | ollama run glm-ocr:latest |
| RX 7600 8GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~64 tok/s | ollama run glm-ocr:latest |
| RTX 3080 10GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~169 tok/s | ollama run glm-ocr:latest |
| M1 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~20 tok/s | ollama run glm-ocr:latest |
| M2 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| M3 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| M4 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~36 tok/s | ollama run glm-ocr:latest |
| RTX 3060 12GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~80 tok/s | ollama run glm-ocr:latest |
| RTX 4070 12GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~112 tok/s | ollama run glm-ocr:latest |
| RTX 5070 12GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~149 tok/s | ollama run glm-ocr:latest |
| RTX 4060 Ti 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~64 tok/s | ollama run glm-ocr:latest |
| RTX 4070 Ti Super 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~149 tok/s | ollama run glm-ocr:latest |
| RTX 4080 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~159 tok/s | ollama run glm-ocr:latest |
| RTX 5080 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~213 tok/s | ollama run glm-ocr:latest |
| RX 7800 XT 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~139 tok/s | ollama run glm-ocr:latest |
| RX 9070 XT 16GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~142 tok/s | ollama run glm-ocr:latest |
| M2 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| M3 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~30 tok/s | ollama run glm-ocr:latest |
| M4 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~36 tok/s | ollama run glm-ocr:latest |
| RTX 3090 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~208 tok/s | ollama run glm-ocr:latest |
| RTX 4090 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~224 tok/s | ollama run glm-ocr:latest |
| RX 7900 XTX 24GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~213 tok/s | ollama run glm-ocr:latest |
| M1 Pro 32GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~59 tok/s | ollama run glm-ocr:latest |
| M2 Pro 32GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~59 tok/s | ollama run glm-ocr:latest |
| M4 32GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~36 tok/s | ollama run glm-ocr:latest |
| M3 Pro 36GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~44 tok/s | ollama run glm-ocr:latest |
| Mac Studio M4 Max (32c GPU) 36GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~121 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Max (32c GPU) 36GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~136 tok/s | ollama run glm-ocr:latest |
| RTX 5090 32GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~398 tok/s | ollama run glm-ocr:latest |
| M4 Pro 48GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~81 tok/s | ollama run glm-ocr:latest |
| Mac Studio M4 Max (40c GPU) 48GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~162 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Max (40c GPU) 48GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~182 tok/s | ollama run glm-ocr:latest |
| M1 Max 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~119 tok/s | ollama run glm-ocr:latest |
| M2 Max 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~119 tok/s | ollama run glm-ocr:latest |
| M3 Max 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~119 tok/s | ollama run glm-ocr:latest |
| M4 Max 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~162 tok/s | ollama run glm-ocr:latest |
| Mac Studio M4 Max (40c GPU) 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~162 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Max (40c GPU) 64GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~182 tok/s | ollama run glm-ocr:latest |
| DGX Spark 64GB (ASUS, Dell, HP, MSI, Acer, Gigabyte) | Ollama | BF16 | 3.7 GB | fits in fast memory | ~76 tok/s | ollama run glm-ocr:latest |
| Mac Studio M3 Ultra 96GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~243 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Ultra 96GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~356 tok/s | ollama run glm-ocr:latest |
| M3 Max 128GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~119 tok/s | ollama run glm-ocr:latest |
| M4 Max 128GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~162 tok/s | ollama run glm-ocr:latest |
| Mac Studio M4 Max (40c GPU) 128GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~162 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Max (40c GPU) 128GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~182 tok/s | ollama run glm-ocr:latest |
| DGX Spark 128GB (NVIDIA, ASUS, Dell, Lenovo, HP, MSI, Acer, Gigabyte) | Ollama | BF16 | 3.7 GB | fits in fast memory | ~76 tok/s | ollama run glm-ocr:latest |
| M2 Ultra 192GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~237 tok/s | ollama run glm-ocr:latest |
| 2x Mac Studio M5 Max (40c GPU) 128GB | MLX | MLX 8bit | 2.6 GB | fits in fast memory | ~435 tok/s | exo # on every Mac (brew install --cask exo); load mlx-community/GLM-OCR-8bit with Tensor sharding + RDMA |
| Mac Studio M3 Ultra 256GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~243 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Ultra (64c GPU) 256GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~356 tok/s | ollama run glm-ocr:latest |
| Mac Studio M5 Ultra (80c GPU) 256GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~356 tok/s | ollama run glm-ocr:latest |
| 2x DGX Spark 128GB | vLLM | Official weights | 3.7 GB | fits in fast memory | ~97 tok/s | vllm serve zai-org/GLM-OCR --tensor-parallel-size 2 --distributed-executor-backend ray --max-model-len 8192 |
| 4x Mac Studio M5 Max (40c GPU) 128GB | MLX | MLX 8bit | 2.6 GB | fits in fast memory | ~684 tok/s | exo # on every Mac (brew install --cask exo); load mlx-community/GLM-OCR-8bit with Tensor sharding + RDMA |
| 2x Mac Studio M5 Ultra (80c GPU) 256GB | MLX | MLX 8bit | 2.6 GB | fits in fast memory | ~851 tok/s | exo # on every Mac (brew install --cask exo); load mlx-community/GLM-OCR-8bit with Tensor sharding + RDMA |
| Mac Studio M3 Ultra (80c GPU) 512GB | Ollama | BF16 | 3.7 GB | fits in fast memory | ~243 tok/s | ollama run glm-ocr:latest |
| 4x DGX Spark 128GB | vLLM | Official weights | 3.7 GB | fits in fast memory | ~139 tok/s | vllm serve zai-org/GLM-OCR --tensor-parallel-size 4 --distributed-executor-backend ray --max-model-len 8192 |
| 4x Mac Studio M5 Ultra (80c GPU) 256GB | MLX | MLX 8bit | 2.6 GB | fits in fast memory | ~1337 tok/s | exo # on every Mac (brew install --cask exo); load mlx-community/GLM-OCR-8bit with Tensor sharding + RDMA |
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-18.
A project by OnDevice Labs · llms.txt