BestLLMfor EN Your hardware. Your LLM. Your call.
APIOpen data Find my LLM
Model fiche

Gemma 3 27B

By Google · United States

Updated 2026-07-13

chat general vision multilingual
Parameters
27B
License
Gemma
Context
125k
VRAM (Q4)
16 GB
Released
March 2025

Overview

Google's flagship Gemma 3 at 27B — multimodal, 128K context, and an LMArena Elo of 1338 that beats Llama 3.1 405B at 15x smaller. Sets the bar for open chat under 30B.

When to pick this model

  • You want the strongest open chat model under 30B
  • You need multimodal input and long context in one model
  • You're standardizing on Google's open stack
  • You have a 24GB+ GPU and want frontier-class chat locally

VRAM requirements by quantization

VRAM REQUIRED (GB)81216243248Q4_K_M16 GBQ5_K_M19 GBQ8_029 GBFP1654 GB
QuantizationVRAM required
Q4_K_M (recommended)16 GB
Q5_K_M19 GB
Q8_029 GB
FP16 (no quantization)54 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, Gemma 3 27B needs a 16 GB card at Q4_K_M (16 GB). Stepping up to Q8_0 nearly doubles the footprint to 29 GB, and unquantized FP16 weights take 54 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.

Without a GPU, Gemma 3 27B needs roughly 28 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 13 tokens/sec on a mid-range card, and up to 32 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.

What hardware do you need

The table below matches Gemma 3 27B 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 memoryExample cardsBest fit for Gemma 3 27B
8 GBRTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GBDoes not fit — needs 16 GB at Q4_K_M
12 GBRTX 5070, RTX 5070 Ti Laptop, RTX 4080 LaptopDoes not fit — needs 16 GB at Q4_K_M
16 GBRTX 5080, RTX 4080 Super, Radeon RX 9070 XTQ4_K_M (16 GB used)
24 GBRTX 4090, Radeon RX 7900 XTX, RTX 5090 LaptopQ5_K_M (19 GB used)
32 GBRTX 5090Q8_0 (29 GB used)

Which GPU should you buy to run Gemma 3 27B?

To run Gemma 3 27B locally at Q4, you need ~16 GB of VRAM. The best value for this is a RTX 5070 Ti (16 GB VRAM).

Check RTX 5070 Ti price on Amazon →

As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.

Published benchmark scores

BenchmarkScore
LMArena Elo73
MMLU78.6
MMLU-Pro67.5
MATH89

Scores published by the model author or aggregated from public leaderboards. Re-measured monthly by our editorial team.

To put Gemma 3 27B in context: its MMLU score of 78.6 ranks #11 of the 34 catalog models with a published MMLU result (catalog median 73.4). Scores are author-published and measured under different harnesses, so treat gaps of a few points as noise rather than a verdict.

Strengths

  • LMArena Elo 1338 — beats Llama 3.1 405B at 15x smaller
  • Multimodal with vision input
  • 128K context window
  • 140 language coverage

Limitations

  • Gemma License rather than Apache
  • No thinking mode for hard reasoning
  • Trails dedicated reasoners on math benchmarks

Typical workloads

In our catalog grid, Gemma 3 27B is filed under Writing, Analysis, Vision — 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 125k-token context window covers long chats and mid-sized documents, though very large retrieval workloads will need chunking. It ships under the Gemma license — commercial use is generally possible but read the specific terms before embedding it in a product.

Architecture & training

Architecture: Dense VLM · sliding-window attention

Training: 14T tokens.

Verdict

Punch-above-its-weight open chat that quietly outscores models 15x its size.

Quick start

ollama run gemma3:27b

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.

Similar models worth comparing

Frequently asked questions

How much VRAM does Gemma 3 27B need?

At the recommended Q4_K_M quantization, Gemma 3 27B needs about 16 GB of VRAM. Q8_0 takes 29 GB, and unquantized FP16 weights take 54 GB.

Can Gemma 3 27B run without a GPU?

Yes — with roughly 28 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 Gemma 3 27B support?

Gemma 3 27B supports a 125k-token context window (128,000 tokens).

Can I use Gemma 3 27B commercially?

Gemma 3 27B ships under the Gemma license. Commercial use is generally permitted subject to its terms — review the license text before shipping a product.

How fast is Gemma 3 27B on consumer hardware?

Our compatibility engine estimates on the order of 13 tokens/sec on a mid-range GPU and up to 32 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.

Which quantization of Gemma 3 27B should I download first?

Start with Q4_K_M (16 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.

Tools

Is Gemma 3 27B the right pick for you?

Compute self-hosted ROI → Back to catalog