CogVLM2 19B needs ~18.4 GB VRAM. MacBook Pro M4 32GB has 23.0 GB. With Q4_K_M quantization, expect ~8 tok/s.
Operating mode
Interactive favors responsiveness, while light API and scale-out lean harder on serving readiness. The fit stays the same, but the recommendation lens changes.
Current mode
Balanced
Balanced for general local use. Keeps the ranking neutral across personal and serving workflows.
Select quantization to explore
Fit status
Runs well
Decode
10.0 tok/s
TTFT
19327 ms
Safe context
8K
Memory
18.4 GB / 23.0 GB
The model fits in shared memory, but shared-memory bandwidth is now the real limiter.
Fit does not mean dedicated-VRAM speed
Unified or shared memory can make a model technically fit, but sustained tokens per second may still trail a discrete high-bandwidth GPU with less total memory.
Shared-memory contention still exists
The OS, browser, and inference runtime all compete for the same physical memory pool, so real-world headroom is less forgiving than raw capacity suggests.
Prioritize bandwidth, not only capacity
If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.
| Workload | Grade | Fit | Decode | TTFT | Context |
|---|---|---|---|---|---|
| Chat | A | Runs well | 10.0 tok/s | 10542 ms | 8K |
| Coding | A | Runs well | 7.5 tok/s | 25970 ms | 8K |
| Agentic Coding | A | Tight fit | 10.0 tok/s | 28111 ms | 8K |
| Reasoning | A | Runs well | 10.0 tok/s | 22841 ms | 8K |
| RAG | A | Tight fit | 10.0 tok/s | 35139 ms | 8K |
How CogVLM2 19B (19B params) fits at each quantization level on MacBook Pro M4 32GB (23.0 GB usable).
| Quant | Bits | VRAM | Quality | Fit |
|---|---|---|---|---|
Q2_K | 2 | 7.4 GB | Low | A81 |
Q3_K_S | 3 | 9.3 GB | Low | A82 |
NVFP4 | 4 |
Copy-paste commands to run CogVLM2 19B on your machine.
Run
docker run --rm -it ghcr.io/ggerganov/llama.cpp:full \
--hf-repo "THUDM/cogvlm2-llama3-chat-19B" \
--hf-file "cogvlm2-llama3-chat-19B-Q4_K_M.gguf" \
-c 4096 -ngl 99Your hardware
| Model | Params | Grade | Decode | Capabilities |
|---|---|---|---|---|
 Qwen3-Coder 30B A3B Instruct | 30.5B | A | 11.7 tok/s | |
| Qwen 3.5 27B | 27B | S | 8.6 tok/s |
Yes, MacBook Pro M4 32GB can run CogVLM2 19B with a A grade (Runs well). Expected decode speed: 7.5 tok/s.
CogVLM2 19B (19B parameters) requires approximately 18.4 GB of memory with Q4_K_M quantization.
The recommended quantization for CogVLM2 19B is Q4_K_M, which balances quality and memory efficiency.
On MacBook Pro M4 32GB, CogVLM2 19B achieves approximately 7.5 tokens per second decode speed with a time-to-first-token of 25970ms using Q4_K_M quantization.
For coding workloads, CogVLM2 19B on MacBook Pro M4 32GB receives a A grade with 7.5 tok/s and 8K context.
On MacBook Pro M4 32GB, CogVLM2 19B can safely use up to 8K tokens of context. The model's official context limit is 8K, but available memory constrains the safe maximum.
Paste this snippet into any page to show a live fit card.
<iframe src="https://willitrunai.com/embed/cogvlm2-19b-on-m4-32gb" width="400" height="180" frameborder="0" style="border:none;border-radius:12px;overflow:hidden;" title="Will It Run AI — fit result"></iframe>
Preview:
10.6 GB |
| Medium |
| A83 |
Q4_K_M | 4 | 11.6 GB | Medium | A84 |
Q5_K_M | 5 | 13.7 GB | High | A83 |
Q6_KBest for your GPU | 6 | 15.6 GB | High | A83 |
Q8_0 | 8 | 20.3 GB | Very High | F0 |
F16 | 16 | 38.9 GB | Maximum | F0 |
| Qwen 3.6 27B | 27B | S | 7.1 tok/s |
| Qwen3-VL 30B A3B Instruct | 30B | S | 12.4 tok/s |
| Qwen 3.5 35B A3B | 35B | A | 10.2 tok/s |
Prioritize bandwidth, not only capacity. If this workload feels slow, the next useful step is often a GPU tier with materially faster memory bandwidth rather than only a small bump in capacity.
Not always. MacBook Pro M4 32GB can often fit larger models thanks to unified memory, but a discrete GPU with dedicated high-bandwidth VRAM may still decode faster once the model fits. For this combination, the important distinction is capacity versus sustained throughput.