utMeta · Downloadable
Llama-3.1-8B-Instruct
Generates text for conversations and writing.
- Source listed
- Jul 18, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3.1
YOUR MEMORY BUDGET
Local AI models for 8GB VRAM.
Explore models by theoretical weight storage. This shortlist covers available downloadable language models with a known or name-derived parameter count.
6 GBleft for weights at 4-bit
8 GB budget − 2 GB reserve
8 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.
852 models within the weight budget
Largest parameter count first · not a quality rankingMeta · Downloadable
Llama-Guard-3-8B
Generates text for conversations and writing.
- Source listed
- Jul 22, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3.1
Meta · Downloadable
Meta-Llama-3-8B
Generates text for conversations and writing.
- Source listed
- Apr 17, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3
Meta · Downloadable
Meta-Llama-3-8B-Instruct
Generates text for conversations and writing.
- Source listed
- Apr 17, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3
Meta · Downloadable
Meta-Llama-Guard-2-8B
Generates text for conversations and writing.
- Source listed
- Apr 17, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3
Microsoft · Downloadable
AesCode-8B
Designed for coding and software tasks. Understands text and images and responds in text.
- Source listed
- Sep 29, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
Microsoft · Downloadable
HARC-Llama-3.1-8B-Instruct
Generates text for conversations and writing.
- Source listed
- Jul 2, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- llama3.1
Microsoft · Downloadable
UserLM-8b
Generates text for conversations and writing.
- Source listed
- Sep 30, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- mit
Mistral AI · Downloadable
Ministral-8B-Instruct-2410
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
- Oct 15, 2024
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- other
NVIDIA · Downloadable
4D-RGPT-8B
Generates text for conversations and writing.
- Source listed
- Jun 2, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- cc-by-nc-4.0
NVIDIA · Downloadable
EGM-8B-SFT
Understands text and images and responds in text.
- Source listed
- Apr 2, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
NVIDIA · Downloadable
Nemotron-Labs-Diffusion-8B
Generates text for conversations and writing.
- Source listed
- Mar 18, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- other
NVIDIA · Downloadable
Nemotron-Labs-Diffusion-VLM-8B
Understands text and images and responds in text.
- Source listed
- May 8, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- other
OpenBMB · Downloadable
BitCPM-CANN-8B
Generates text for conversations and writing.
- Source listed
- May 15, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MathForm-8B
Generates text for conversations and writing.
- Source listed
- Aug 14, 2026
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B
Generates text for conversations and writing.
- Source listed
- Jun 5, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-Eagle-FRSpec
Generates text for conversations and writing.
- Source listed
- Jun 5, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-Eagle-FRSpec-QAT-cpmcu
Generates text for conversations and writing.
- Source listed
- Jun 5, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-Eagle-vLLM
Generates text for conversations and writing.
- Source listed
- Jun 6, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-marlin-cpmcu
Generates text for conversations and writing.
- Source listed
- Jun 6, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-marlin-Eagle-vLLM
Generates text for conversations and writing.
- Source listed
- Jun 6, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4-8B-marlin-vLLM
Generates text for conversations and writing.
- Source listed
- Jun 6, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4.1-8B
Generates text for conversations and writing.
- Source listed
- Sep 2, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
OpenBMB · Downloadable
MiniCPM4.1-8B-AutoAWQ
Generates text for conversations and writing.
- Source listed
- Sep 4, 2025
- Parameters
- 8B
- 4-bit weights only
- ≈ 4 GB
- License
- apache-2.0
What changes with a 8GB budget?
With 2 GB set aside, 6 GB remains for weights. At 4-bit that allows up to 12 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