@sualehasif996: We've gotten really really good at RL. Composer 2.5 is fighting well-above its weight class. Very excited for the next …
Summary
Cursor introduces Composer 2.5, an improved AI model that is more intelligent, better at sustained work, and more reliable at following complex instructions. The release highlights advancements in reinforcement learning and scaling.
View Cached Full Text
Cached at: 05/19/26, 12:40 AM
We’ve gotten really really good at RL. Composer 2.5 is fighting well-above its weight class.
Very excited for the next release as we scale model sizes and FLOPs with @SpaceXAI!
Cursor (@cursor_ai): Introducing Composer 2.5, our most powerful model yet.
It’s more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions.
For the next week, we’re doubling the included usage of the model.
Similar Articles
@cursor_ai: Introducing Composer 2.5, our most powerful model yet. It's more intelligent, better at sustained work on long-running …
Cursor AI announces Composer 2.5, their most powerful model yet, with enhanced intelligence, better sustained work on long tasks, and improved instruction following; they are doubling included usage for a week.
Composer 2.5
Cursor released Composer 2.5, its most powerful AI model yet, aiming to improve code generation capabilities.
Cursor Introduces Composer 2.5
Cursor released Composer 2.5, a major update to its AI coding assistant featuring improved intelligence, behavior, and training via targeted reinforcement learning and increased compute, built on Moonshot's Kimi K2.5.
@elonmusk: Try it out! (Partially trained on Colossus 2)
Cursor announces Composer 2.5, an upgraded AI model for coding that is more intelligent and reliable, partially trained on Colossus 2. Usage is doubled for the next week.
@srush_nlp: Talk: Training Composer https://youtube.com/watch?v=uTgqYeVxy2c… Overview of the methods that we use at Cursor to build…
Cursor shared the training methods for its self-developed programming model Composer 2, including large-scale continuous pre-training, long-range reinforcement learning, and an internal benchmark CursorBench, which brings the model's programming performance to a top level.