← CanIRunLocal · All hardware · All models
Mistral · 1050B parameters (52B active) · chat, reasoning, coding, vision.
Ollama lists mistral-large-4:cloud as a cloud model. Open weights are announced but not yet downloadable, so there is no local setup command.
Ollama lists this as a cloud model (ollama run mistral-large-4:cloud). The library page describes a multimodal Mixture-of-Experts model with 52B active of 1.05T total parameters and a 1M context; the readme also says 49B active of 1T. Open weights were announced for 2026-10-31 and are not on Hugging Face yet, so there is no local download.
No listed GPU, Mac, or multi-node cluster has enough fast memory for the verified weights.
None of the listed machines have enough fast memory.
Estimates, not benchmarks. Model sizes are ollama.com download sizes; memory needed adds headroom for ~8k tokens of context; speeds are rough figures from memory bandwidth. Data updated 2026-10-20.
A project by OnDevice Labs · llms.txt