Gemma 3 27B IT
27BGoogle Gemma 3
Gemma 3 large instruct with long context and multimodal support. Q4 ~16GB — dual-GPU or 24GB card with short ctx.
131K
Max Context
3
Quant Variants
GGUF Q4_K_M
Best Quality
+2.8%
Perplexity rise (best build)
Quantization Variants
Per-quant VRAM, quality loss, and inference speed on RTX 4090
Measured = run on this site · Estimated = not run here: calculated, or a figure this site has not verified · Community = public reports
| Format | Level | BPW | VRAM | PPL Loss | Speed | Source | Actions |
|---|---|---|---|---|---|---|---|
| GGUF | Q4_K_M | 4.85 | 18.4 GB | 2.8% | 48 tok/s | Community | |
| GGUF | Q3_K_M | 3.87 | 14.9 GB | 5.5% | 55 tok/s | Estimated | |
| AWQ | INT4 | 4 | 15.3 GB | 3.8% | 62 tok/s | Estimated |
Which file each size is for
A file size is not a memory figure: the VRAM column is the calculator's total at 4K context — weights plus KV cache plus a 10% activation buffer. Units differ on purpose: file sizes are decimal GB, as Hugging Face lists them; the VRAM column is GiB (1024³ bytes) with 2% added to the weights, so 7.12 GB of file counts as about 6.8 GiB there. A vision projector (mmproj), where the repo ships one, is a separate download and is in neither number.
- GGUF Q4_K_MSized at the calculator's generic rate for this level; no file was checked.
- GGUF Q3_K_MSized at the calculator's generic rate for this level; no file was checked.
- AWQ INT4Sized at the calculator's generic rate for this level; no file was checked.
Similar models
Compare with Gemma 3Gemma 3 4B IT
Google Gemma 3
Google Gemma 3 multimodal 4B. 128K context; strong vision + text on 8GB cards.
Gemma 3 12B IT
Google Gemma 3
Mid-size Gemma 3 with vision. Fits 16GB at Q4; excellent multilingual chat.
Qwen3-VL 30B-A3B Instruct
Alibaba Qwen3-VL
Multimodal MoE with only ~3B active parameters, so it stays responsive on Apple unified memory and survives CPU offload far better than a dense 30B. Q4 ~19GB — a 24GB card holds it outright.
Gemma 4 26B-A4B IT
Google Gemma 4
Google's Gemma 4 mixture-of-experts model: about 25B parameters in the text model, roughly 4B active per token, a 256K window and image input. Only 5 of its 30 layers keep a cache that grows with context, and those use fewer, wider heads, so the cache stays small — about 0.9 GB at 32K and 5.3 GB at the full 256K. At Q4_K_M that is about 18.8 GB at 32K: comfortable on a 24 GB card, which can still load the whole window (tight, about 23.6 GB). A 16 GB card cannot hold it, since the Q4_K_M file alone is 17.0 GB. The vision encoder is a separate 1.2 GB file on top. Sizes come from bartowski's GGUF files, whose Q4_K_M works out at 5.40 bits per weight, well above the generic 4-bit rate; Unsloth's UD-Q4_K_M is a different file, 0.09 GB smaller. No per-level quality loss has been published.
Running Gemma 3 27B IT locally
At Q4_K_M and 4K of context, Gemma 3 27B IT needs about 18.4 GB — 15.8 GB of weights, 0.92 GB of KV cache and a 1.7 GB activation buffer. The smallest cards in this index that clear that comfortably are 24 GB ones — 4 of them, from the RTX 3090 up — and 31 of the 78 cards here do. These are calculated figures, not measurements: the estimate stops counting a card as comfortable at 88% of its VRAM, which is roughly the room a desktop session needs.
What longer context costs
Going from 4K to 32K adds about 2.19 GB, taking the total to 20.8 GB. That is gentler than the parameter count suggests: only 10 of 62 layers keep a KV cache that grows with context, and 52 more cap theirs at a sliding window, so the cache scales at a fraction of the usual rate. The model's native window is 128K; holding all of it at Q4_K_M would need about 29 GB.
Which build to download
This index tracks 2 formats for it — GGUF, AWQ — across 3 levels. Q4_K_M carries the lowest published perplexity loss at 2.8%. The fastest level listed is AWQ INT4 at about 62 tok/s on an RTX 4090, batch 1 — an estimate, not a run on this site's hardware. GGUF runs on llama.cpp and Ollama across NVIDIA, AMD and Apple silicon; AWQ and GPTQ target vLLM on CUDA and ROCm; EXL2 is ExLlamaV2 — CUDA only, and now archived, so local use rather than serving.
Common questions
- How much VRAM does Gemma 3 27B IT need?
- About 18.4 GB at Q4_K_M with 4K of context and batch 1, rising to roughly 20.8 GB at 32K. That figure is weights plus KV cache plus a 10% activation buffer, calculated from the model's architecture rather than measured on a card.
- Will Gemma 3 27B IT run on a 24GB GPU?
- Yes — at Q4_K_M and 4K context it needs about 18.4 GB, which leaves 5.6 GB spare on a RTX 3090, or on any of the other 24 GB cards here (4 in all). That is the smallest size in this index that clears it comfortably; 31 of 78 do.
- Which quantization of Gemma 3 27B IT should I use?
- Q4_K_M has the lowest published quality loss (2.8%), and Q4_K_M is the level most people run. All 3 levels in the index are Q4_K_M, Q3_K_M, AWQ INT4.
Where this model fits
Sized at Q4_K_M with a 4K context window, smallest card first. Comfortable means the estimate uses at most 88% of the memory.
- Runs comfortably on
- RTX 309024GB · +5.6Radeon RX 7900 XTX24GB · +5.6RTX 409024GB · +5.6Tesla P40 24G24GB · +5.6Mac M5 32G32GB · +5.6
- Tight but possible
- Radeon RX 7900 XT20GBMac M4 24G24GBMac M4 Pro 24G24GB
- Compared here
- GGUF vs AWQ
- Try it yourself
- Size it in the calculatorGenerate the run command
How to actually run this
Deployment guides for this model and this class of hardware.
This page's figures change when the model or the runtime does.Last updated 2026-10-28 RSS → /feed.xml