M4 24GB vs NVIDIA RTX 5080 for local LLMs
8 reports on the M4 24GB and 26 on the NVIDIA RTX 5080
Which is faster for local LLMs?
On Llama 3.2 3B at 4-bit, the M4 24GB has 46.0 generation tokens/s (one run); prompt processing 530.0 tokens/s (one run); the NVIDIA RTX 5080 has 298.6 generation tokens/s (one run); prompt processing 19397.4 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 8.0x higher on the NVIDIA RTX 5080 (120 GB/s on the M4 24GB, 960 on the NVIDIA RTX 5080).
- Memory
- 24GB unified
- Memory bandwidth
- 120 GB/s
- FP16 compute
- 4.3 TFLOPS
- Reports
- 8
- Memory
- 16GB
- Memory bandwidth
- 960 GB/s
- FP16 compute
- 225 TFLOPS
- Reports
- 26
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 | M4 24GB | NVIDIA RTX 5080 |
|---|---|---|
| Llama 3.2 3B4-bit | 46.0one runPP 530.0 (one run)llama.cppcontext not stated | 298.6one runPP 19397.4 (one run)Ollama16K 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 | M4 24GB | NVIDIA RTX 5080 |
|---|---|---|
| Llama 3.2 3B | 15.2 t/sFP16 | 88.0 t/sFP16 |
| Qwen3.8 27B | 10.8 t/sQ2_K | 62.6 t/sQ2_K |
| Qwen3.8 125B · 6B active | doesn't fit at these settings | doesn't fit at these settings |
| Qwen3.6 35B · 3B active | 35.8 t/sQ2_K | 207.0 t/sQ2_K |
| DeepSeek V4 Flash 284B · 13B active | doesn't fit at these settings | doesn't fit at these settings |
| Qwen3.6 27B | 10.8 t/sQ2_K | 62.6 t/sQ2_K |
| Gemma 4 26B · 4B active | 24.1 t/sQ3_K_S | 139.6 t/sQ3_K_S |
Change the context, RAM or model in the calculator: M4 24GB · NVIDIA RTX 5080
Every report on each card, with its source: M4 24GB · NVIDIA RTX 5080