Infrastructure / Deployment

Ollama

Local LLM runtime that wraps llama.cpp in a user-friendly CLI and REST API with a Docker-like model management experience (ollama pull, ollama run). Handles model downloading, quantization selection, and GPU/CPU routing automatically. 172K GitHub stars, the most popular way for developers to run models locally. Became the default local inference backend for dozens of AI applications and MCP integrations.

Active 2026; ~170K+ GitHub stars, library of thousands of model builds (GGUF)

Openness

5 high confidence

5.0

license: MIT
source: public
core-gated: ungated

license:MIT;source:public;no-feature-gated-core

https://github.com/ollama/ollama/blob/main/LICENSE recorded 2026-06-04
flagship phase-C verification source
https://github.com/ollama/ollama recorded 2026-06-04
flagship phase-C verification source
https://ollama.com/ recorded 2026-06-04
flagship phase-C verification source

Adoption

4 high confidence

4.0

Reported 52M+ monthly model pulls in early 2026; 170K+ GitHub stars. Pull/usage volume places it in the 1-10M+ active-user range; rated 4 conservatively on monthly pull volume (not stars).

https://github.com/ollama/ollama recorded 2026-06-04
flagship phase-C verification source
https://www.ollama.com/ recorded 2026-06-04
flagship phase-C verification source

Capability

4 high confidence

4.0

Local model serving with one-command run, OpenAI-compatible REST API, broad model library (Llama, Qwen, Gemma, gpt-oss, DeepSeek). Serving-layer tool; no MLPerf result cited, rated on feature/coverage matrix.

https://github.com/ollama/ollama recorded 2026-06-04
flagship phase-C verification source
https://ollama.com/library recorded 2026-06-04
flagship phase-C verification source

Unchanged since 2026-07-30 (last edited, not re-checked)