EXAONE 4.5 33B
LG AI Research's multimodal Korean flagship: a 33B model with 256k context that lands in the top 10 of the Artificial Analysis Intelligence Index.
By LG AI Research · South Korea
Updated 2026-09-15
When to pick this model
- Korean-language production workloads
- Bilingual EN/KR multimodal applications
- Vision tasks needing a compact 33B footprint
- Long-context multimodal analysis
VRAM requirements by quantization
| Quantization | VRAM required |
|---|---|
| Q4_K_M (recommended) | 20 GB |
| Q5_K_M | 24 GB |
| Q8_0 | 36 GB |
| FP16 (no quantization) | 66 GB |
VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.
In practice, EXAONE 4.5 33B wants a 24 GB card at Q4_K_M (20 GB). Stepping up to Q8_0 nearly doubles the footprint to 36 GB, and unquantized FP16 weights take 66 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.
Without a GPU, EXAONE 4.5 33B needs roughly 32 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 3 tokens/sec on entry-level GPUs, on the order of 12 tokens/sec on a mid-range card, and up to 30 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.
What hardware do you need
The table below matches EXAONE 4.5 33B to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.
| GPU memory | Example cards | Best fit for EXAONE 4.5 33B |
|---|---|---|
| 8 GB | RTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GB | Does not fit — needs 20 GB at Q4_K_M |
| 12 GB | RTX 5070, RTX 5070 Ti Laptop, RTX 4080 Laptop | Does not fit — needs 20 GB at Q4_K_M |
| 16 GB | RTX 5080, RTX 4080 Super, Radeon RX 9070 XT | Does not fit — needs 20 GB at Q4_K_M |
| 24 GB | RTX 4090, Radeon RX 7900 XTX, RTX 5090 Laptop | Q5_K_M (24 GB used) |
| 32 GB | RTX 5090 | Q5_K_M (24 GB used) |
Which hardware should you buy to run EXAONE 4.5 33B?
To run EXAONE 4.5 33B locally at Q4, you need ~20 GB for Q4 weights alone. Hardware option to compare: GMKtec EVO-X2 64GB / 1TB (Ryzen AI Max+ 395). Leave memory for the system and context; verify inference-engine support. A mini PC does not provide CUDA or macOS/MLX.
As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.
Strengths
- 262k context in a 33B model
- Integrated vision capabilities at this scale
- Strong Korean and English performance
- Top-10 placement on independent intelligence benchmarks
Limitations
- Around 20 GB VRAM at Q4
- EXAONE license requires review for commercial use
- Smaller English-focused community than Llama or Qwen
Typical workloads
In our catalog grid, EXAONE 4.5 33B is filed under Advanced Korean, Multimodal, Asian Multilingual — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: vision-language work — screenshots, charts, scanned documents; multilingual workloads.
The 256k-token context window is large enough to hold entire codebases' worth of files or long reports in a single prompt, which is what makes local RAG and document analysis practical. It ships under the EXAONE AI Model License license — commercial use is generally possible but read the specific terms before embedding it in a product.
Architecture & training
Architecture: Dense · 33B · EXAONE 4.5 · LG AI Research · integrated vision
Training: LG AI Research — Korean+English corpus, multimodal vision added, 262k ctx.
The clear pick for Korean multimodal work — capable, compact, and competitive globally, with licensing caveats to verify.
Quick start
Install the runtime for your system: Windows, macOS or Linux. Check the exact model tag or GGUF quantization below; catalog IDs are not necessarily Ollama tags.
Start at 4096 tokens of context, then use ollama ps to check GPU/CPU placement. A default download may use a different quantization from the configurator’s memory estimate. Keep the free setup working before considering a kit.
ollama run exaone4.5:33bOr use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.
Your private ChatGPT, free, on your own machine in an hour — LM Studio, Ollama, Open WebUI, your documents, no cloud.
- Lifetime online access
- PDF + files
- 30-day refund
Similar models worth comparing
Frequently asked questions
How much VRAM does EXAONE 4.5 33B need?
At the recommended Q4_K_M quantization, EXAONE 4.5 33B needs about 20 GB of VRAM. Q8_0 takes 36 GB, and unquantized FP16 weights take 66 GB.
Can EXAONE 4.5 33B run without a GPU?
Yes — with roughly 32 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.
What context window does EXAONE 4.5 33B support?
EXAONE 4.5 33B supports a 256k-token context window (262,144 tokens).
Can I use EXAONE 4.5 33B commercially?
EXAONE 4.5 33B ships under the EXAONE AI Model License license. Commercial use is generally permitted subject to its terms — review the license text before shipping a product.
How fast is EXAONE 4.5 33B on consumer hardware?
Our compatibility engine estimates on the order of 12 tokens/sec on a mid-range GPU and up to 30 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.
Which quantization of EXAONE 4.5 33B should I download first?
Start with Q4_K_M (20 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. On a 24 GB card you can run up to Q5_K_M.