Can I Run It?
Check which open-weights AI models your workstation or laptop hardware can run locally across GGUF quantization levels.
YOUR HARDWARE
DeepSeek R1 32B
32B reasoning model distilled from Qwen 32B with R1 chain-of-thought
SmolLM3 3B
HuggingFaceLightweight multilingual reasoning model for edge hardware
Llama 3.2 3B
MetaLightweight Llama model optimized for mobile and edge devices
Qwen 3.5 2B
AlibabaSmall multimodal Qwen 3.5 model with ultra-fast generation
DeepSeek R1 1.5B
DeepSeekCompact reasoning model distilled from DeepSeek-R1
Qwen 3.5 9B
AlibabaMultimodal Qwen 3.5 mid-size model for general reasoning and code
Llama 3.1 8B
MetaState-of-the-art 8B open model for chat, coding, and function calling
Qwen 2.5 Coder 7B
AlibabaSpecialized code generation model with impressive benchmark performance
DeepSeek R1 14B
DeepSeekMid-size reasoning distilled model with chain-of-thought capabilities
Gemma 2 9B
GoogleGoogle's 9B open model with high math and reasoning density
Mistral 7B v0.3
Mistral AIClassic open-weights model with function calling and sliding window attention
Phi-3.5 Mini 3.8B
MicrosoftHighly optimized 3.8B model for long-context comprehension and logic
Qwen 2.5 Coder 32B
AlibabaFrontier 32B code model matching Sonnet 3.5 in coding benchmarks
DeepSeek R1 32B
DeepSeek32B reasoning model distilled from Qwen 32B with R1 chain-of-thought
Mixtral 8x7B 47B
Mistral AISparse Mixture-of-Experts architecture (8x7B params, 13B active)
Llama 3.3 70B
MetaMeta's flagship 70B open weight model for enterprise agentic intelligence
DeepSeek R1 70B
DeepSeek70B reasoning powerhouse distilled from Llama 70B