A number of Large Language Models are distributed as open-source or open-weights.
And the gap in performance between open and proprietary models is closing:
https://www.reddit.com/r/LocalLLaMA/comments/1dzrjn2/open_llms_catching_up_to_closed_llms_codingelo/
Benchmarks
https://www.vellum.ai/open-llm-leaderboard
Models by year
Bonsai 27B (2006)
https://prismml.com/news/bonsai-27b
Ternary logic version of Qwen 3.6 27B Runs on phones and laptops
GLM 5.2 (2026)
https://ollama.com/library/glm-5.2
https://www.wafer.ai/blog/glm52-amd
Swiss AI Apertus (2026)
Liquid.ai LFM2 (2026)
https://www.liquid.ai/blog/lfm2-24b-a2b
OpenAI Privacy Filter (2026)
https://openai.com/index/introducing-openai-privacy-filter/
Ternary Bonsai (2026)
https://prismml.com/news/ternary-bonsai
Gemma 4 (2026)
https://ollama.com/library/gemma4
https://deepmind.google/models/gemma/gemma-4/
MiniMax M2 (2026)
https://ollama.com/library/minimax-m2.7
https://www.minimax.io/news/minimax-m27-en
DeepSeek
https://ollama.com/library/deepseek-v3.2
Architecture
https://arcprize.org/blog/r1-zero-r1-results-analysis
Source
https://github.com/deepseek-ai/DeepSeek-R1
Qwen 3
https://ollama.com/library/qwen3.5
https://quesma.com/blog/qwen-36-is-awesome/
Kimi K2 (2025)
https://composio.dev/content/notes-on-kimi-k2
https://intuitionlabs.ai/articles/kimi-k2-technical-deep-dive
01.ai Yi
https://github.com/01-ai/Yi-1.5
Meta Llama 3 (2024)
https://github.com/meta-llama/llama3
Mistral 7B
https://www.secondstate.io/articles/mistral-7b-instruct-v0.1/
Fine-tune Mistral models
https://github.com/mistralai/mistral-finetune
- LoRa
- single-node, multi-GPU