Written by Jakub Rusinowski · Last updated February 10, 2026
Yes — comfortably
Yes, comfortably — EXAONE 3.5 2.4B at Q8_0 needs about 4 GB of the 6.4 GB usable on 8 GB system RAM, leaving ~2.4 GB spare and running at ~22.3 tok/s (estimated), with room for about 32,768 tokens of context.
Confidence: high · Recommended quantization: Q8_0 · Estimated speed: ~22.3 tok/s
| Usable memory for models | 6.4 GB |
| Memory bandwidth | 90 GB/s |
| Quant | Memory needed | Fits 6.4 GB? | Max context | Est. speed | Download |
|---|---|---|---|---|---|
| F16 | 6.2 GB | ✓ Yes | 8K | ~12.8 tok/s | 4.8 GB |
| Q8_0 | 4 GB | ✓ Yes | 32K | ~22.3 tok/s | 2.6 GB |
| Q6_K | 3.4 GB | ✓ Yes | 32K | ~27.6 tok/s | 2 GB |
| Q5_K_M | 3.1 GB | ✓ Yes | 32K | ~31 tok/s | 1.7 GB |
| Q4_K_M | 2.9 GB | ✓ Yes | 32K | ~35.1 tok/s | 1.4 GB |
| Q3_K_M | 2.5 GB | ✓ Yes | 32K | ~45.1 tok/s | 1 GB |
| Q2_K | 2.2 GB | ✓ Yes | 32K | ~53.6 tok/s | 0.8 GB |
| Variant | Needs at Q4_K_M | Fits? | Est. speed |
|---|---|---|---|
| EXAONE 3.5 32B | 22.3 GB | ✗ Too large | — |
| EXAONE 3.5 7.8B | 6.6 GB | ✗ Too large | — |
| EXAONE 3.5 2.4B | 2.9 GB | ✓ Fits | ~35.1 tok/s |
llama.cpp (CPU build) or Ollama — both run without a GPU
Yes, comfortably — EXAONE 3.5 2.4B at Q8_0 needs about 4 GB of the 6.4 GB usable on 8 GB system RAM, leaving ~2.4 GB spare and running at ~22.3 tok/s (estimated), with room for about 32,768 tokens of context.
Q8_0 — it needs about 4 GB of the 6.4 GB available, downloads as roughly 2.6 GB, and runs at an estimated 22.3 tokens/sec with up to 32K of context.
Nothing binding — the model fits with headroom and generates at a usable speed on this hardware.
llama.cpp (CPU build) or Ollama — both run without a GPU
← Can I Run It? | EXAONE 3.5 model page | Check your hardware