Gemma 3 27B
Google's Gemma 3 model card lists a 27B size. This page uses that size class, not a separate counted total.
The Gemma 3 model card says 128k context for the 4B, 12B, and 27B models.
From the model card
Strengths
- The Gemma 3 card describes a multimodal model: text and image in, text out.
- The card says 128k context and coverage of 140+ languages. 27B is the largest of the three sizes on that card.
Weaknesses
- The license is the Gemma terms and the prohibited-use policy, not Apache-2.0.
- The text KV formula does not include the vision encoder.
- No Frankenstein row yet.
Sources: Gemma 3 model card · config.json (public mirror; official weights are gated)
Memory at 8,192 tokens
Computed, not measured. Weights use llama.cpp’s whole-file averages (Q4_K_M = 4.8944 bits/weight, Q8_0 = 8.5008) on the total parameter count. KV is fp16 and does not depend on the quant. The formula, including the sliding-window and latent-KV cases, is on the methodology page.
| Q4_K_M | Q8_0 | |
|---|---|---|
| Weights | 15.38 GiB | 26.72 GiB |
| KV at 8,192 | 1.03 GiB | 1.03 GiB |
| Total | 16.42 GiB Computed | 27.75 GiB Computed |
62 layers, 16 KV heads, head dim 128. Sliding window 1024. A layer is full attention when (layer index + 1) is divisible by 6; the others use the window.
Tokens per second
Not yet measured. This lab has no Frankenstein row for Gemma 3 27B, so there is no estimated tok/s either.
Hardware tiers
Minimum is the smallest RAM that holds Q4_K_M plus 8k KV. Suggested does not name a second machine or a speed this lab has not run. Performance means a CUDA-class card with 12 GB or more of GDDR6 or GDDR6X and the weights fully on the card. The Tesla P40 is 24 GB of GDDR5, so a P40 speed is not the performance-tier number.
| Tier | Example hardware | Quant | Fit | tok/s |
|---|---|---|---|---|
| Minimum | CPU-only, with at least 18.42 GiB of system RAM for the Q4_K_M weights, 8k KV, and a 2 GiB planning allowance. | Q4_K_M | CPU-only, if the RAM is there | CPU: not yet measured |
| Suggested | A machine you are willing to wait on, or a GPU that can hold the quant. This lab has not measured a separate “comfortable” machine, so this row does not invent one. | Q4_K_M until a Q8 run exists | Whichever of the machines above actually holds the weights | not yet measured |
| Performance | A CUDA-class card with enough GDDR6 or GDDR6X to hold about 17.42 GiB (Q4_K_M weights, 8k KV, and a 1 GiB planning allowance). 12 GB is not enough for that total. | Q4_K_M | Full VRAM only if the card is larger than 12 GB | not yet measured |
Where it sits
Best-for is a judgment from the card’s stated jobs and the model’s size. It is not a speed ranking. The matching build is Used 24 GB GPU. Score it against a machine you have on What can my rig run?