LG AI Research · exaone
Can I run exaone-deep?
EXAONE Deep exhibits superior capabilities in various reasoning tasks including math and coding benchmarks, ranging from 2.4B to 32B parameters developed and released by LG AI Research.
7.8B parameters32K contexttextEXAONE AI Model License Agreement 1.1 - NC9 builds
exaone-deep system requirements
Memory needed at 4K context, counting model weights, KV cache, compute buffers and runtime overhead. Download sizes come from the official Ollama registry.
| Build | Quantization | Download | Memory needed |
|---|---|---|---|
| exaone-deep:2.4b | Q4_K_M | 1.5 GB | ~2.6 GB(est.) |
| exaone-deep:2.4b-q8_0 | Q8_0 | 2.6 GB | ~4.2 GB(est.) |
| exaone-deep:7.8b | Q4_K_M | 4.4 GB | ~6.7 GB(est.) |
| exaone-deep:2.4b-fp16 | F16 | 5.0 GB | ~7.5 GB(est.) |
| exaone-deep:7.8b-q8_0 | Q8_0 | 7.7 GB | ~11.5 GB(est.) |
| exaone-deep:7.8b-fp16 | F16 | 14.6 GB | ~21.2 GB(est.) |
| exaone-deep:32b | Q4_K_M | 18.0 GB | ~26.2 GB(est.) |
| exaone-deep:32b-q8_0 | Q8_0 | 31.7 GB | ~45.7 GB(est.) |
Figures marked (est.) approximate the KV cache because this model's architecture details are not published. They are indicative rather than exact.
How to run exaone-deep locally
- 1.Install Ollama for Windows, macOS or Linux.
- 2.Run this in a terminal:
ollama run exaone-deep:2.4b - 3.The weights download on first run, then an interactive prompt opens.
Frequently asked questions about exaone-deep
- How much VRAM do I need to run exaone-deep?
- At 4K context, the smallest published build of exaone-deep (Q4_K_M) needs roughly 2.6 GB of GPU memory once weights, KV cache, compute buffers and runtime overhead are counted. Bigger quantizations and longer context windows need more. It can also run partly or entirely on the CPU using system RAM, more slowly.
- Can I run exaone-deep without a dedicated GPU?
- Yes, but slowly. Without a GPU the model runs on the CPU using system RAM, which usually means a few tokens per second rather than dozens. You would need at least 2.6 GB of free RAM for the smallest build.
- How big is the exaone-deep download?
- The smallest published build is 1.5 GB. There are 9 builds in total, the largest being 59.6 GB. Leave some extra free disk space beyond the download itself.
- How do I run exaone-deep locally?
- Install Ollama, then run "ollama run exaone-deep:2.4b" in a terminal. The weights download on first use and an interactive session opens.
Will exaone-deep run on your PC?
Scan your hardware once and get a verdict for this model and every build — free, no account.
Model data from the official Ollama registry · verified 13/09/2026