NVIDIA L4 vs M5 Ultra 256GB for local LLMs
3 reports on the NVIDIA L4 and 24 on the M5 Ultra 256GB
Which is faster for local LLMs?
On Qwen3.8 27B at 4-bit, the NVIDIA L4 has 14.6 generation tokens/s (one run); the M5 Ultra 256GB has 50.0 generation tokens/s (one run); prompt processing 1800.0 tokens/s (one run). Those are community reports whose engines and settings differ, so the gap is a rough sign, not a controlled measure of the two cards. Memory bandwidth, which caps how fast a card can write, is 4.1x higher on the M5 Ultra 256GB (300 GB/s on the NVIDIA L4, 1,229 on the M5 Ultra 256GB).
- Memory
- 24GB
- Memory bandwidth
- 300 GB/s
- FP16 compute
- 121 TFLOPS
- Reports
- 3
- Memory
- 256GB unified
- Memory bandwidth
- 1,229 GB/s
- FP16 compute
- not on record
- Reports
- 24
Measured on both
Generation and prompt processing speeds from plain runs (one device, one request, no speculative decoding, the whole model in the device's memory) of the same model size at the same bit class. Each side shows the engines and reported context behind its median. Context can mean a configured window or prompt depth, so these are community comparisons with differing settings. How to read these
| Model | NVIDIA L4 | M5 Ultra 256GB |
|---|---|---|
| Qwen3.8 27B4-bit | 14.6one runllama.cppcontext not stated | 50.0one runPP 1800.0 (one run)oMLX8K context |
Estimated by the calculator
The quant the calculator recommends for each card and its estimated speed at a 32,768-token context with 32 GB of system RAM, from memory bandwidth and the model's shape. Where a model was also measured above, the measurement wins.
| Model | NVIDIA L4 | M5 Ultra 256GB |
|---|---|---|
| Qwen3.8 27B | 10.9 t/sQ4_K_M | 14.9 t/sFP16 |
| Qwen3.8 125B · 6B active | 16.4 t/sQ2_K, experts in RAM | 50.8 t/sQ8_0 |
| Qwen3.6 35B · 3B active | 38.5 t/sQ4_K_S | 53.5 t/sFP16 |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | 66.9 t/sQ4_K_M |
| Qwen3.6 27B | 10.9 t/sQ4_K_M | 14.9 t/sFP16 |
| Gemma 4 26B · 4B active | 24.5 t/sQ5_K_S | 40.1 t/sFP16 |
| GLM-5.3 320B · 18B active | doesn't fit at these settings | 35.6 t/sQ3_K_M |
Change the context, RAM or model in the calculator: NVIDIA L4 · M5 Ultra 256GB
Every report on each card, with its source: NVIDIA L4 · M5 Ultra 256GB