Gemma 3 12B IT
The 12B Gemma 3 instruction model adds a larger text-and-image family option; this estimate covers text weights only.
- Provider
- Family
- Gemma 3
- Parameters
- 12B
- Formats
- gguf, safetensors
Curated model catalog
Browse local language models
Compare cataloged model families, parameter sizes, and recorded formats. Open a model for its quantization candidates, sources, and compatibility notes.
20 models found.
Gemma 3 1B IT
Google's compact instruction-tuned Gemma 3 text model for constrained local systems.
- Provider
- Family
- Gemma 3
- Parameters
- 1B
- Formats
- gguf, safetensors
Gemma 3 27B IT
Google's largest Gemma 3 instruction variant, useful for exploring high-memory local configurations.
- Provider
- Family
- Gemma 3
- Parameters
- 27B
- Formats
- gguf, safetensors
Gemma 3 4B IT
Google's 4B instruction-tuned Gemma 3 model with text and image-input capabilities in its published model family.
- Provider
- Family
- Gemma 3
- Parameters
- 4B
- Formats
- gguf, safetensors
Llama 3.1 8B Instruct
Meta's 8B instruction-tuned Llama 3.1 model with publisher-documented long-context metadata.
- Provider
- Meta
- Family
- Llama 3.1
- Parameters
- 8B
- Formats
- gguf, safetensors
Llama 3.2 1B Instruct
Meta's compact instruction-tuned Llama model for lightweight local text generation.
- Provider
- Meta
- Family
- Llama 3.2
- Parameters
- 1B
- Formats
- gguf, safetensors
Llama 3.2 3B Instruct
Meta's 3B instruction-tuned Llama model, suitable for modest local hardware when quantized.
- Provider
- Meta
- Family
- Llama 3.2
- Parameters
- 3B
- Formats
- gguf, safetensors
Llama 3.3 70B Instruct
Meta's 70B text-only instruction model, extending the catalog into the high-memory class.
- Provider
- Meta
- Family
- Llama 3.3
- Parameters
- 70B
- Formats
- gguf, safetensors
Mistral 7B Instruct v0.3
Mistral's 7B instruction-tuned model with a broad ecosystem of local GGUF conversions.
- Provider
- Mistral AI
- Family
- Mistral
- Parameters
- 7B
- Formats
- gguf, safetensors
Mistral Nemo 12B Instruct
A 12B Mistral and NVIDIA collaboration model, extending the catalog into a larger multilingual instruction tier.
- Provider
- Mistral AI / NVIDIA
- Family
- Mistral Nemo
- Parameters
- 12B
- Formats
- gguf, safetensors
Mistral Small 3.2 24B Instruct
Mistral's 24B instruction-tuned model; the memory estimate covers text weights and not its image-processing path.
- Provider
- Mistral AI
- Family
- Mistral Small 3.2
- Parameters
- 24B
- Formats
- gguf, safetensors
Phi-3.5 Mini Instruct
Microsoft's compact Phi instruction model with a documented long-context capability.
- Provider
- Microsoft
- Family
- Phi-3.5
- Parameters
- 3.8B
- Formats
- gguf, safetensors
Phi-4 Mini Instruct
Microsoft's 3.8B instruction model provides a compact alternative with a publisher-documented 128K context limit.
- Provider
- Microsoft
- Family
- Phi-4
- Parameters
- 3.8B
- Formats
- gguf, safetensors
Qwen2.5 3B Instruct
Qwen's 3B instruction-tuned model with a published 32K context configuration for its GGUF variant.
- Provider
- Qwen
- Family
- Qwen2.5
- Parameters
- 3B
- Formats
- gguf, safetensors
Qwen2.5 7B Instruct
Qwen's 7B instruction-tuned model for users with more memory and a need for a broader general-purpose model.
- Provider
- Qwen
- Family
- Qwen2.5
- Parameters
- 7B
- Formats
- gguf, safetensors
Qwen2.5-Coder 7B Instruct
A code-focused instruction model, adding a distinct software-development use case to the catalog.
- Provider
- Qwen
- Family
- Qwen2.5-Coder
- Parameters
- 7.6B
- Formats
- gguf, safetensors
Qwen3 14B
Qwen's 14.8B general-purpose Qwen3 model, filling the size range between the existing 8B and 27B entries.
- Provider
- Qwen
- Family
- Qwen3
- Parameters
- 14.8B
- Formats
- gguf, safetensors
Qwen3 32B
Qwen's 32.8B general-purpose Qwen3 model for users evaluating the larger end of the local model range.
- Provider
- Qwen
- Family
- Qwen3
- Parameters
- 32.8B
- Formats
- gguf, safetensors
Qwen3 4B
Qwen's compact 4B general-purpose model; a separate GGUF conversion is listed for llama.cpp workflows.
- Provider
- Qwen
- Family
- Qwen3
- Parameters
- 4B
- Formats
- gguf, safetensors
Qwen3 8B
An 8B Qwen3 general-purpose model that adds a mid-sized option beyond the existing Qwen2.5 entries.
- Provider
- Qwen
- Family
- Qwen3
- Parameters
- 8B
- Formats
- gguf, safetensors