Skip to main content
kitlocal.dev

Ollama Model Reference

Runs locally

100% processed in your browser. Inputs and outputs never leave your device.

Browse Ollama model database with sizes, capabilities, and usage info

llama3.1

Meta's latest Llama model. Great general-purpose performance.

Params: 8B/70B/405BSize: 4.7GB/40GB/231GBContext: 128K
generalcodechat
$ ollama run llama3.1

llama3.2

Lightweight Llama models optimized for edge and mobile.

Params: 1B/3BSize: 1.3GB/2.0GBContext: 128K
edgesmall
$ ollama run llama3.2

mistral

Mistral 7B - efficient model with strong performance.

Params: 7BSize: 4.1GBContext: 32K
generalfast
$ ollama run mistral

mixtral

Mixture of Experts model. Sparse activation for efficiency.

Params: 8x7B/8x22BSize: 26GB/80GBContext: 64K
generalMoE
$ ollama run mixtral

codellama

Code-specialized Llama. Supports code completion and infilling.

Params: 7B/13B/34B/70BSize: 3.8GB/7.4GB/19GB/40GBContext: 16K
codecompletion
$ ollama run codellama

deepseek-coder-v2

DeepSeek Coder v2. State-of-the-art coding model.

Params: 16B/236BSize: 8.9GB/133GBContext: 128K
codemath
$ ollama run deepseek-coder-v2

phi3

Microsoft Phi-3. Small but capable reasoning model.

Params: 3.8B/14BSize: 2.2GB/7.9GBContext: 128K
reasoningsmall
$ ollama run phi3

gemma2

Google Gemma 2. Efficient open model family.

Params: 2B/9B/27BSize: 1.6GB/5.4GB/16GBContext: 8K
generalgoogle
$ ollama run gemma2

qwen2.5

Alibaba Qwen 2.5. Strong multilingual and coding.

Params: 0.5B-72BSize: 0.4-41GBContext: 128K
multilingualcode
$ ollama run qwen2.5

nomic-embed-text

Text embedding model for RAG and semantic search.

Params: 137MSize: 274MBContext: 8K
embeddingRAG
$ ollama run nomic-embed-text