Cursor has been a trusted partner of Anthropic since Sonnet 3.5. We’ll continue to increase compute to support Claude models in Cursor and are excited for what comes next with them at SpaceX.
We’re sorry to see that OpenAI put out a note saying they plan to block Cursor users from accessing OpenAI models in three months.
OpenAI models serve about 5% of Cursor user traffic, and we’re speaking with the OpenAI team to resolve this.
Cursor was one of the very first users of OpenAI, we’ve worked closely with their team for years, and we’ve trusted their platform to be neutral infrastructure for our business.
Show more
The 1.5T that could!
Bigger and even better models on the horizon.
Introducing Grok 4.6.
It delivers frontier intelligence and is a significant improvement over Grok 4.5 at the same price.
Each wave of ai product birthed the dominant player of that era.
There was OpenAI and chat, then Anthropic and coding.
Welcome SpaceXAI and general knowledge work.
Our user base in India has tripled in the past year, with more agent requests per user than any other country.
Cursor Start gives generous access to Grok 4.5 and Composer for ₹649/month. Both are extremely capable models while remaining token efficient and cheap to use.
Excited for India's builders to try it!
Show more
Today we're launching Cursor Start, a new ₹649/month plan for developers in India.
Start includes generous access to Grok 4.5 and Composer, so you can plan, build, test, and ship with agents every day.
Show more
Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.
Show more
Very excited for this model. It’s an enormous improvement over composer 2.5 and trained entirely from scratch. It’s been a pleasure working with the SpaceXAI team on it.
Even more excited by the slope of the effort and the models to come.
Show more
We've partnered with SpaceXAI to train Grok 4.5.
It’s our most powerful model yet and the first we've built for more than software engineering.
Excited to train some very strong models!
SpaceX has exercised the option to acquire @cursor_ai in an all-stock transaction with the goal of building the world’s most useful AI models.
For the past few months, SpaceXAI has been jointly training a model with Cursor, which will be released in Cursor and Grok Build soon.
We look forward to working closely with the Cursor team to advance our frontier AI capabilities
Show more
It's a good model
We're releasing a technical report describing how Composer 2 was trained.
We're increasing capacity for Composer 2, 2x more usage all weekend! Try it out!
We've evaluated a lot of base models on perplexity-based evals and Kimi k2.5 proved to be the strongest!
After that, we do continued pre-training and high-compute RL (a 4x scale-up).
The combination of the strong base, CPT and RL, and Fireworks' inference and RL samplers make Composer-2 frontier level.
It was a miss to not mention the Kimi base in our blog from the start. We'll fix that for the next model.
Show more
Congrats to the @cursor_ai team on the launch of Composer 2!
We are proud to see Kimi-k2.5 provide the foundation. Seeing our model integrated effectively through Cursor's continued pretraining & high-compute RL training is the open model ecosystem we love to support.
Note: Cursor accesses Kimi-k2.5 via @FireworksAI_HQ ' hosted RL and inference platform as part of an authorized commercial partnership.
Show more
RT @Kimi_Moonshot: Congrats to the @cursor_ai team on the launch of Composer 2!
We are proud to see Kimi-k2.5 provide the foundation. S…
It's a really good model! Excited for more people to try it
Composer 2 marks the one-year anniversary of our large model training efforts. Since then, we've built an exceptionally talent-dense team of ~40 people with some of the best researchers and engineers from the labs, academia, industry, and more heterogeneous backgrounds.
And we are exclusively focused on coding. We don't care about models that can respond to emails, do your tax returns, or be your friend.
Every FLOP, token, parameter, and researcher is entirely dedicated to software engineering.
Show more
Writing software is becoming effortless.
So much has changed and yet there's still so much to go. Cloud agents and demo-based review are a big leap toward this future.
The future of review is not code. It's agents proving their work.
For products this is demo videos.
And we're slowly discovering what this looks like for the rest of software engineering!
Cursor now shows you demos, not diffs.
Agents can use the software they build and send you videos of their work.
We trained the best coding model in the world under 1T parameters.
Excited for people to try it!
Composer 1.5 is now available.
We’ve found it to strike a strong balance between intelligence and speed.
Very excited about this.
Rebuilding the primitives of software engineering in an agent-first world!
today we introduced Cursor Blame
it turns out it's useful to persist the "why" behind your code
so your teammates can understand code lineage
...and future agents can use this context to make better decisions
Show more
When does synchronous coding hit diminishing returns to intelligence?
I believe we're a few months away from a ChatGPT intelligence bar where >95% of these queries don't get much better with smarter models.
More intelligence would only matter for async-level tasks, ones that would take developers hours to perform.
Think about a large fraction of UI work. The bottleneck for layout changes is not intelligence or task horizon, it's user intent.
In that world, the speed matters a lot more! I'm very excited for frontier models at Composer-1 speed.
Show more
By flushing chat transcripts to disk, Cursor can recall information across a conversation millions of tokens long.
A step function better for infinite-length conversations and long-horizon tasks.
Show more
Cursor's agent now uses dynamic context for all models.
It's more intelligent about how context is filled while maintaining the same quality. This reduces total tokens by 46.9% when using multiple MCP servers.
Show more
