Laya — local AI model by ConvAI Innovations
Written by Jakub Rusinowski · Last updated
Small encoder-based decision models (ModernBERT / mmBERT) for fast routing and guardrails in 100+ languages, installed with pip.
Variants
The smallest Laya variant needs about 1 GB of VRAM at Q4_K_M — quantized weights plus framework overhead, before any KV cache.
| Model | VRAM at Q4 | VRAM | Context | Run it |
|---|---|---|---|---|
| Laya → 421M / 322M | ~1.1 GB | 512 tokens | laya-serve / Python Router |
Memory is quantized weights plus overhead at Q4_K_M, from the same engine as the GPU & VRAM checker.
How to run Laya locally
Install Ollama, then pull the tag.
Served by laya-serve / Python Router. This model does not run in Ollama.
Pick a size above for its own VRAM figure, speed estimate and install command.
Licence
Commercial use permitted. No usage restrictions beyond attribution.
Applies to: LayaRecommended GPU
The cheapest catalogued GPU that runs Laya locally (min 1 GB VRAM) is the Intel Arc B570 (10 GB).
Laya — frequently asked questions
How much VRAM does Laya need?
Laya needs about 1 GB VRAM at Q4_K_M quantization for its smallest variant. Variants: Laya (1 GB, Q4_K_M). On Apple Silicon, unified memory counts toward this requirement.
Can I run Laya on an RTX 4090 (24 GB)?
Yes — Laya runs on an RTX 4090 (24 GB) and other 24 GB cards such as the RTX 3090. Smaller variants also fit comfortably on 8–16 GB GPUs at Q4_K_M.
What quantization should I use for Laya?
Q4_K_M is the best balance of quality and VRAM for Laya in most cases. Choose Q8_0 for near-lossless quality if you have spare VRAM, or smaller quants (Q3/Q2) only when memory is tight.
How do I run Laya locally?
Laya is a decision model: it is called over an HTTP endpoint (/v1/systemone), not chatted with. Where a variant is available in Ollama 0.35 or newer, pull it with `ollama pull` and send requests to the local server; the others ship their own server. Each variant page shows the exact commands for that model.