YOUR MEMORY BUDGET

Local AI models for 16GB VRAM.

Explore models by theoretical weight storage. This shortlist covers available downloadable language models with a known or name-derived parameter count.

14 GBleft for weights at 4-bit
16 GB budget − 2 GB reserve

The reserve is your assumption, not a measured overhead allowance. The default 2 GB may be too little for your runtime or context length. Results are not confirmed fits and have no verified speed rating. Weight estimates use decimal GB and do not include quantization overhead. For timed results on one 12GB machine, see measured speeds.

1,031 models within the weight budget

Largest parameter count first · not a quality ranking

JetBrains · Downloadable

deepseek-coder-6.7B-kexer

Text generationCoding

Designed for coding and software tasks. Generates text for conversations and writing.

Source listed
May 13, 2024
Parameters
6.7B
4-bit weights only
≈ 3.4 GB
License
apache-2.0
Compare this model →

Microsoft · Downloadable

wavecoder-ultra-6.7b

Text generationCoding

Designed for coding and software tasks. Generates text for conversations and writing.

Source listed
Apr 11, 2024
Parameters
6.7B
4-bit weights only
≈ 3.4 GB
License
mit
Compare this model →

Z.ai · Downloadable

LongAlign-6B-64k

EmbeddingsText generation

Generates text for conversations and writing.

Source listed
Jan 29, 2024
Parameters
6B
4-bit weights only
≈ 3 GB
License
apache-2.0
Compare this model →

Google · Downloadable

gemma-3n-E2B

VisionVideo understandingTranscription

Understands text and images and responds in text.

Source listed
Jun 12, 2025
Parameters
5.4B
4-bit weights only
≈ 2.7 GB
License
gemma
Compare this model →

Google · Downloadable

gemma-3n-E2B-it

VisionVideo understandingTranscription

Understands text and images and responds in text.

Source listed
Jun 12, 2025
Parameters
5.4B
4-bit weights only
≈ 2.7 GB
License
gemma
Compare this model →

Meta · Downloadable

MobileMoE-L-Base

Text generation

Generates text for conversations and writing.

Source listed
Jun 10, 2026
Parameters
5.3B
4-bit weights only
≈ 2.7 GB
License
fair-noncommercial-research-license
Compare this model →

Meta · Downloadable

MobileMoE-L-SFT

Text generation

Generates text for conversations and writing.

Source listed
Jun 11, 2026
Parameters
5.3B
4-bit weights only
≈ 2.7 GB
License
fair-noncommercial-research-license
Compare this model →

NVIDIA · Downloadable

NV-Reason-CT

VisionReasoning

Understands text and images and responds in text. Supports a reasoning mode.

Source listed
Sep 8, 2026
Parameters
5.3B
4-bit weights only
≈ 2.7 GB
License
openmdw-1.1
Compare this model →

Datalab · Downloadable

chandra-ocr-2

VisionDocument reading

Understands text and images and responds in text.

Source listed
Mar 16, 2026
Parameters
5.3B
4-bit weights only
≈ 2.6 GB
License
openrail
Compare this model →

OpenBMB · Downloadable

cpm-bee-5b

EmbeddingsText generation

Generates text for conversations and writing.

Source listed
May 27, 2023
Parameters
5B
4-bit weights only
≈ 2.5 GB
License
Not disclosed
Compare this model →

What changes with a 16GB budget?

With 2 GB set aside, 14 GB remains for weights. At 4-bit that allows up to 28 billion parameters in the theoretical calculation. An 8-billion-parameter model alone needs approximately 4 GB at this precision, before its conversation cache and runtime.

Why might a listed model still fail to load?

Its actual weight format may need more space, the runtime may not support its architecture, or the conversation cache may exceed your reserve. Vision models can also need space for image processing. Check the original model card and measure your exact setup before relying on it.

Read the memory guide and technical sources · Compare two models