Gemma-SEA-LION-v4-27B-IT
Gemma-SEA-LION-v4-27B-IT is an open-weights language model from AI Singapore, released August 20, 2025. It has 27B parameters (dense).
With SEA-LION v4, we introduce our first multimodal model, built on Gemma 3 (27B). It is our most powerful and efficient, multimodal & multilingual model yet, designed to run even on a laptop.
Check it against your own PC →
Lab
AI Singapore
Released
August 20, 2025
Weights
Open
Parameters
27B dense
Context
–
Price / M tokens
–
Epoch ECI
not yet rated
Can you run Gemma-SEA-LION-v4-27B-IT locally?
The smallest common GPU it runs on is 12 GB of VRAM (with 32 GB of RAM): IQ3_XXS, split GPU / CPU, roughly 7 tokens per second. Estimates assume 32 GB of DDR5 RAM, a 16K context and llama.cpp-style GGUF quantization. Enter your exact hardware in the live app.
| GPU memory | Verdict | Best quant | VRAM | RAM | tok/s |
|---|---|---|---|---|---|
| 6 GB VRAM | Very slow | IQ3_XXS (noticeable loss) | 5.2 GB | 8.9 GB | ~3 |
| 8 GB VRAM | Very slow | IQ3_XXS (noticeable loss) | 7.2 GB | 6.9 GB | ~4 |
| 12 GB VRAM | Runs, slowly | IQ3_XXS (noticeable loss) | 11.2 GB | 2.9 GB | ~7 |
| 16 GB VRAM | Runs on GPU | IQ3_XXS (noticeable loss) | 14.1 GB | – | ~28 |
| 24 GB VRAM | Runs on GPU | Q5_K_M (excellent) | 23.0 GB | – | ~28 |
| 32 GB VRAM | Runs on GPU | Q6_K (≈ lossless) | 25.9 GB | – | ~46 |
Benchmark results
Epoch AI has not published results for Gemma-SEA-LION-v4-27B-IT yet. New models are usually evaluated within days to weeks.