Organisations

African AI company building language technology and models for African languages.

Alibaba Cloud team publishing the Qwen language and multimodal model families.

AI research company publishing the DeepSeek open-weight language and reasoning model families.

AI company publishing the GLM language-model family and related open model research.

Research arm of MediaTek publishing model research relevant to efficient and on-device AI.

Samsung's advanced research organisation working on on-device AI and language-model technology.

Shanghai AI Laboratory team publishing the InternLM open model family and associated deployment and agent tooling.

Ant Group AI research team publishing open language, reasoning, and multimodal model families.

AI company publishing the Kimi open-weight language, reasoning, and coding model families.

Singapore national AI programme publishing SEA-LION models for Southeast Asian languages and cultures.

Abu Dhabi research institute publishing the Falcon open model family.

French AI company publishing Mistral and Mixtral open-weight models.

Generative AI company publishing open-weight image, language, audio, and video models.

Nonprofit research lab publishing open models and datasets, including the multilingual Aya programme.

Nonprofit AI research institute publishing open models, datasets, and evaluations, including OLMo.

MIT research lab developing efficient AI methods including AWQ quantization and TinyChat deployment.

GPU vendor publishing open models, inference runtimes, quantization tooling, and optimized model artifacts.

CPU, GPU, and NPU vendor providing ROCm and Ryzen AI software for local model inference.

CPU, GPU, NPU, and accelerator vendor developing OpenVINO for local and edge inference.

Mobile and PC silicon vendor providing Qualcomm AI Hub models and on-device inference tooling.

Open engineering consortium maintaining MLPerf training, inference, and client benchmarks.

Open research organisation developing model systems and evaluation projects including Chatbot Arena and FastChat.

Platform hosting models, datasets, Spaces, libraries, and community evaluation projects used throughout local and open AI.

Technology company publishing the Gemma family of open models and on-device AI tooling.

Technology company publishing the Phi open model family and local inference tooling including ONNX Runtime.

Meta research organisation publishing the Llama open-weight model family and related open AI research.

AI company publishing the Liquid Foundation Model family for efficient on-device and edge deployment.

IBM team publishing the Granite family of open language, code, and enterprise models.