Really interesting release from Mistral. Good to see a relatively big dense model. These are hard to come by these days, most labs have moved to mixture-of-experts at this scale.
Context window is 256k, which is on the smaller end compared to recent releases pushing 1M plus, but probably fine for most document processing and reasoning workloads.
Curious how it holds up on long-context and coding tasks against the MoE flagships.
Mistral Vibe@mistralvibe
Mistral Medium 3.5, a new flagship model in public preview by @MistralAI that merges instruction-following, reasoning, and coding into a single 128B dense model with a 256k context window and configurable reasoning effort. It's a new default model for Mistral Vibe and Le Chat. Released as open weights, under a modified MIT license.
English
