Ollama
Ollama is a platform that enables users to run large language models (LLMs) locally on their machines. It simplifies managing open-source LLMs by packaging model weights, configuration, and runtime into a single bundle. Supports REST APIs, Python, and JavaScript SDKs. Models include Llama, Gemma, Qwen, DeepSeek, and more. It requires CPU and system RAM, with GPU acceleration via NVIDIA CUDA or experimental AMD ROCm for optimal performance.
What’s new
Verified 5h agoGemma 4 is now significantly faster in Ollama on Apple Silicon, generating tokens nearly 90% faster on average across a coding-agent benchmark by leveraging multi-token prediction (MTP). Ollama auto-tunes how many tokens to draft as it runs, so the speedup is on by default, requires no configuration, and does not change the model's output.
## What's Changed
- Tightened Gemma 4 MoE model loading in the MLX engine
- Updated the MLX engine to the latest version, including a new small-batch matmul kernel
- Updated the underlying llama.cpp engine to build 9840
- Improved Gemma 4 multi-token prediction (MTP) performance
**Full Changelog**: https://github.com/ollama/ollama/compare/v0.30.12...v0.31.1
Official download sources
Each source below has been validated against an approved publisher domain and editorially approved. Links route through a safe internal redirect.
Platforms: Linux, macOS, Windows
Ollama compared
A factual, side-by-side overview based only on verified data.
Alternatives to Ollama
Verified, official downloads in the same category.
Frequently asked questions
Where can I download Ollama safely?
Which platforms does Ollama support?
Is this the official Ollama download?
Who develops Ollama?
What are the best alternatives to Ollama?
Reviews & ratings
No reviews yet — be the first to share your experience.
Write a review