YOUR MEMORY BUDGET

Local AI models for 24GB VRAM.

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

22 GBleft for weights at 4-bit
24 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.

1,129 models within the weight budget

Largest parameter count first · not a quality ranking

PrismML · Downloadable

Bonsai-27B

Text generation

Generates text for conversations and writing.

Source listed
Jul 4, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

Qwen · Alibaba · Downloadable

Qwen3.5-27B

Vision

Understands text and images and responds in text.

Source listed
Feb 24, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

Qwen · Alibaba · Downloadable

Qwen3.6-27B

Vision

Understands text and images and responds in text.

Source listed
Apr 21, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

Qwen · Alibaba · Downloadable

Qwen3.8-27B

Vision

Understands text and images and responds in text.

Source listed
Aug 5, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

Tencent · Downloadable

UI-Mate-27B

Vision

Understands text and images and responds in text.

Source listed
Aug 14, 2026
Parameters
27B
4-bit weights only
≈ 13.5 GB
License
apache-2.0
Compare this model →

Liquid AI · Downloadable

LFM2-24B-A2B

Text generation

Generates text for conversations and writing.

Source listed
Feb 24, 2026
Parameters
24B
4-bit weights only
≈ 12 GB
License
other
Compare this model →

Liquid AI · Downloadable

LFM2.5-24B-A2B

Text generation

Generates text for conversations and writing.

Source listed
Oct 12, 2026
Parameters
24B
4-bit weights only
≈ 12 GB
License
other
Compare this model →

Mistral AI · Downloadable

Mistral-Small-24B-Base-2501

Text generation

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Jan 23, 2025
Parameters
24B
4-bit weights only
≈ 12 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Mistral-Small-24B-Instruct-2501

Text generation

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Jan 28, 2025
Parameters
24B
4-bit weights only
≈ 12 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Devstral-Small-2505

Text generationCoding

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
May 12, 2025
Parameters
23.6B
4-bit weights only
≈ 11.8 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Devstral-Small-2507

Text generationCoding

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Jul 4, 2025
Parameters
23.6B
4-bit weights only
≈ 11.8 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Magistral-Small-2506

Text generation

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Jun 4, 2025
Parameters
23.6B
4-bit weights only
≈ 11.8 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Magistral-Small-2507

Text generation

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Jul 18, 2025
Parameters
23.6B
4-bit weights only
≈ 11.8 GB
License
apache-2.0
Compare this model →

Mistral AI · Downloadable

Mistral-Small-Instruct-2409

Text generation

Generates text, according to its published model architecture (MistralForCausalLM). No task is listed on Hugging Face, so check the model card for intended use.

Source listed
Sep 17, 2024
Parameters
22.2B
4-bit weights only
≈ 11.1 GB
License
other
Compare this model →

What changes with a 24GB budget?

With 2 GB set aside, 22 GB remains for weights. At 4-bit that allows up to 44 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