GPT-OSS
GPT-OSS — OpenAI open-weight language models: gpt-oss-120b for production reasoning and gpt-oss-20b for local use, Apache-2.0
OLMo-core est un projet de LLM locaux open source, créé par allenai (allenai/Olmo-core sur GitHub). Il compte 1 772 étoiles sur GitHub et 364 forks, il est publié sous la licence Apache-2.0 et a été mis à jour pour la dernière fois sur GitHub le 2 novembre 2026.
OLMo-core — the current PyTorch building blocks for the OLMo open language model ecosystem, Apache-2.0
Clone the current OLMo-core repository from GitHub.
Install in editable mode with all extras for training work, or take the published package for library use.
Use the documented modelling and training building blocks for your own OLMo-based model or experiment.
The successor the old OLMo repo itself points to — actively pushed, not a frozen snapshot.
Modelling and training components for the OLMo ecosystem, rather than a single fixed model script.
OLMo is the fully open model family from AI2: open weights, data and code are the stated principles of the project.
API documentation for the library lives on its Read the Docs site.
The README install route: clone, then editable install with all extras.
git clone https://github.com/allenai/Olmo-core.git
cd Olmo-core
pip install -e .[all]The published package for library use.
pip install ai2-olmo-coreModelling and training docs for the OLMo building blocks.
https://olmo-core.readthedocs.io/en/latest/The Allen Institute for AI (AI2), as the building-blocks library for its OLMo model family.
Yes — Apache-2.0 licensed.
No — the old allenai/OLMo repo says it is out of date and no longer active and points here; OLMo-core is the current codebase.
No — it is a PyTorch library of modelling and training building blocks for researchers and engineers.
Clone allenai/Olmo-core, pip install -e .[all], and follow the Read the Docs documentation.
GPT-OSS — OpenAI open-weight language models: gpt-oss-120b for production reasoning and gpt-oss-20b for local use, Apache-2.0
Cloud-native runtime for building and running agentic AI in production.
DeepSeek-Coder-V2 — DeepSeek's open code model family that challenged closed-source code models.
Private AI chat and model studio that runs in your browser via WebGPU.
StepFun's open 196B MoE model that activates 11B parameters per token, built for fast agentic reasoning.
Fine-tune LLMs from one YAML. Layer streaming trains an 8B model on a 4 GB laptop GPU.
Les autres projets de la catégorie LLM locaux avec le plus d'étoiles du catalogue LibreAI — la sélection réaliste d'alternatives si OLMo-core ne vous convient pas.
The simplest way to run open language models on your own machine.
User-friendly AI Interface (Supports Ollama, OpenAI API, ...)
LLM inference in C/C++
A high-throughput and memory-efficient inference and serving engine for LLMs
🤯 LobeHub is your Chief Agent Operator, organizing your agents into 7×24 operations by hiring, scheduling, and reporting on your entire AI team.
Projets de la catégorie LLM locaux proches dans le catalogue, les plus proches en étoiles de OLMo-core — continuez à explorer de part et d'autre.
AI file organizer that sorts and renames files by understanding content.
AI OSINT agent with 20 tools and REPL — for authorized research only.
Elegant AI chat client with workspaces, plugins, and local-first sync.
Whisper-powered subtitle generation: Web UI, CLI, and package.
Local-first agent harness with loop, memory, and evals in readable code.
Self-evolving general AI agent running fully locally, a Manus-style alternative.
Gratuit et ouvert : récupérez le code source, mettez une étoile au dépôt et commencez à créer.