On June 18, 2026, American AI company Poolside released its proprietary model 'Laguna M.1' as an open model, originally released in April 2026 and distributed via API. The model has a 256K context length, and both base and post-trained checkpoints are now available on Hugging Face under Apache 2.0. A quantized model that can be run locally on Macs has also been released. Benchmark results show that 'Laguna M.1 (225B-A23B)' beats the French-made Devstral 2 but scores lower than the Chinese-made DeepSeek-V4-Flash and Qwen 3.5. Among community-released quantization models, 'Laguna-M.1-MLX-Q3' is said to be executable locally on a Mac with an M3 Max processor and 128GB of memory, achieving approximately 26 tok/s with about 100 GB peak memory.