Back to all tools
Ollama
Run LLMs locally — Llama, Mistral, Gemma, and more with one command.
0Open Source
LLM Platforms
Overview
Ollama lets you run large language models locally on your laptop or server with a single command, providing an OpenAI-compatible API for easy integration.
Available Models
— may change as new versions release| Model | Context | Best For | Cost Tier |
|---|---|---|---|
| llama3.3 | 128K | Best all-round open model, general tasks | Free |
| deepseek-r1 | 128K | Reasoning, math, logic with chain-of-thought | Free |
| qwen2.5-coder | 128K | Code generation and completion | Free |
| mistral | 32K | Fast, lightweight general assistant | Free |
| gemma3 | 128K | Google-tuned, efficient on consumer hardware | Free |
| phi4 | 16K | Small but capable, great on CPU/low VRAM | Free |
| nomic-embed-text | 8K | Local text embeddings for RAG pipelines | Free |
Key Features
- One-command model download and run
- OpenAI-compatible REST API on localhost:11434
- Supports Llama 3, Mistral, Gemma, CodeLlama, Phi-3, and 100+ models
- GPU acceleration (NVIDIA, AMD, Apple Silicon)
- Model library with community-contributed models
- Modelfiles for custom model configurations
Real-World Workflows
Run a private AI assistant with no cloud costs
- 1Install Ollama: curl -fsSL https://ollama.com/install.sh | sh
- 2Pull a model: ollama pull llama3
- 3Start chatting: ollama run llama3
- 4Use the local API at http://localhost:11434 from any app
Local AI for coding with Continue IDE extension
- 1Pull a code model: ollama pull codellama or ollama pull deepseek-coder
- 2Install Continue extension in VS Code
- 3Point Continue to Ollama in config.json
- 4Get private, offline code completions and chat
Getting Started
# Install (macOS/Linux)
curl -fsSL https://ollama.com/install.sh | sh
# Pull and run a model
ollama pull llama3
ollama run llama3
# Use the OpenAI-compatible API
curl http://localhost:11434/v1/chat/completions \
-d '{"model":"llama3","messages":[{"role":"user","content":"Hi"}]}'Compare Alternatives
See how Ollama stacks up against similar tools.