Cursor

Cursor released the Environments API in beta, letting developers create and manage cloud agent environments in code, including cloned repos and setup.

@cursor_ai on X

Cursor cloud agents can now build interactive canvases, such as dashboards and reports, which you can share with your team as a link.

@cursor_ai on X

Cursor added Muse Wonder from Meta Superintelligence Labs, priced at $4 per million input tokens and $20 per million output tokens.

@cursor_ai on X

Cursor opened its marketplace to public submissions, so any team making developer tools can now publish an app.

@cursor_ai on X

Cursor's code hosting platform Origin now supports over 20 integrations for CI, deployments, code review, security, monitoring and project management.

@cursor_ai on X

Cursor added Gemini 4 Argon, which scores 52.8% on CursorBench and costs $2 per million input tokens and $10 per million output.

@cursor_ai on X

Cursor is now part of The X Subscription, with shared usage across Grok products, and third-party models in Cursor move to API pricing on new plans.

@cursor_ai on X

Cursor lowered Claude Sonnet 5.5 cache read pricing to $0.10 per million tokens, down from $0.20.

@cursor_ai on X

Cursor added Claude Haiku 5.5, which costs 10x less than Claude Haiku 4.5 on shorter requests and is enabled in Settings.

@cursor_ai on X

Cursor now lets you control agents running on your computer from your phone, using the new Cursor iOS app to check in, reply or start tasks.

@cursor_ai on X

Cursor SDK background subagents now report back, returning results to the parent run as a follow-up turn until every subagent finishes.

@cursor_ai on X

Cursor SDK custom tools can now carry MCP annotations such as readOnlyHint and destructiveHint, so the model can tell a lookup from a delete.

@cursor_ai on X

Cursor SDK agents can now be steered while they run, using run.steer() to add a message to the next turn.

@cursor_ai on X

Cursor SDK users can now replace Cursor's system prompt with their own, with rules, skills and tool schemas still loading, enabled account by account.

@cursor_ai on X

Cursor Rollouts now finds the pull request behind a regression and opens an issue, and one click starts a cloud agent to fix it.

@cursor_ai on X

Cursor added GLM 5.3 and GLM 5.3 Flash, with GLM 5.3 Max the best-scoring open-weight model on CursorBench 4.0.

@cursor_ai on X

Cursor can now build charts and diagrams in chat with /visualize, available in the Agents Window.

@cursor_ai on X

Cursor added Claude Sonnet 5.5, which performs on par with Opus in many tasks.

@cursor_ai on X

Cursor reduced token costs by 7% with no drop in agent quality, through tighter prompts, selective tool loading, better caching and compressed file reads.

@cursor_ai on X

Cursor updated Security Reviewer, which now finishes 21% faster, averaging 3.8 minutes instead of 4.8, for Teams and Enterprise plans.

@cursor_ai on X

Cursor launched Rollouts, which write a monitoring plan and watch changes as they deploy so regressions are caught before users see them.

@cursor_ai on X

Cursor added Claude Opus 5.5, the top model on CursorBench at 57.8% (Max), costing 40% less per task than Opus 5.

@cursor_ai on X

Cursor introduced Projects in beta, where you work with a coordinator agent in one persistent thread that manages subagents.

@cursor_ai on X

Cursor added Muse Spark 1.3 from Meta as an available model.

@cursor_ai on X

Cursor added Gemini 3.8 Flash as an available model.

@cursor_ai on X

Cursor cloud agents can now run on your own infrastructure, including machine pools that scale automatically, while the agent loop stays in Cursor.

@cursor_ai on X

Cursor added Claude Fable 5.1, which scores 73.4% on CursorBench 3.2 at max effort and is strong at verifying its own work.

@cursor_ai on X

Cursor now lets you create new web apps, store the code with Origin and deploy to Vercel.

@cursor_ai on X