Can I run Laguna S 2.1 on an RTX 4090 (24GB)?
Computed with the open FitLLM engine — accurate per-layer KV-cache modeling, not a naive estimate. Updated 2026-08-24.
Memory breakdown (Q4_K_M, F16 KV, 33K context)
| Model weights | 67.0 GB |
| KV cache | 1.6 GB |
| Runtime overhead + reserve | 11.3 GB |
| Total used | 79.8 / 24 GB |
| Short by | 55.8 GB |
Max context that fits at Q4_K_M: does not fit.
Every quantization on the RTX 4090
| Weight quant | Weights | Fits (KV F16) | Used @32K |
|---|---|---|---|
| Q4_K_M | 67.0 GB | ❌ won't fit | 79.8 / 24.0 GB |
| Q5_K_M | 78.1 GB | ❌ won't fit | 92.2 / 24.0 GB |
| Q6_K | 89.8 GB | ❌ won't fit | 105 / 24.0 GB |
| Q8_0 | 116 GB | ❌ won't fit | 135 / 24.0 GB |
| FP16 | 219 GB | ❌ won't fit | 250 / 24.0 GB |
Lower weight quants free memory at some output-quality cost — Q4 is the common sweet spot; below that quality drops faster.
KV cache is F16 here (llama.cpp default). Drop it to Q8/Q4 (-ctk/-ctv) for more context.
Embed this verdict
Live badge for your README or model card — recomputed by the engine, never stale:
[](https://fitllm.run/can-i-run/laguna-s-2-1-on-rtx-4090)
Or from your terminal (exit 0/1 — works as a pre-download guard):
npx fitllm "Laguna S 2.1" --gpu "RTX 4090"Why most VRAM calculators get this wrong
Laguna S 2.1 interleaves sliding-window (local) and global attention 5:1. The local layers cap their KV cache at the 512-token window, and the global layers use a different head shape (head_dim 128 vs 128). A naive "all layers × full context × one head_dim" formula over-counts KV cache by several times.
What fits on the RTX 4090 instead
same GPU Models that fit on the RTX 4090: GLM-4.7-Flash, gpt-oss-20b, Qwen 3.6 27B, Gemma 4 e2b, Gemma 4 e4b, Gemma 4 12b, Gemma 4 26b A4B, Llama-3.2-3B-Instruct, Llama-3.1-8B-Instruct, MiniCPM5-1B, Qwen3-0.6B, Qwen3-1.7B, Llama-3.2-1B-Instruct, Gemma-3-1B-it, Qwen 3.8 27B.
same model GPUs that run Laguna S 2.1: RTX PRO 6000 Blackwell (96GB), 4× RTX 3090 (96GB), H200 141GB (141GB), B200 (180GB).
Reproduce it
Laguna S 2.1 = 117.562B (8B active, MoE), 48 layers. The RTX 4090 has 24GB / 1008GB/s. Same math, open source: fitllm-engine. GGUF bpw from llama.cpp.
All numbers are computed by the open-source fitllm-engine (MIT) from official model config.json values — reproduce or audit them yourself. Estimates; real usage varies with runtime (llama.cpp / MLX / Ollama), driver and display. Found a mismatch? Report it. · FitLLM home