Cursor Launches Composer 2.5: Better Long-Running Agent Work, New Pricing Tiers, and 2× Included Usage This Week
Cursor has released Composer 2.5, calling it its most powerful Composer model yet. Cursor says it’s more intelligent, better at sustained work on long-running tasks, and more reliable at following complex instructions. Cursor’s launch post also says included usage is doubled for the first week. Cursor’s blog post adds token pricing for Standard vs Fast modes.
Cursor announced Composer 2.5 as a new release in its Composer model line, positioning it as the company’s most capable coding model so far. The update matters because Cursor is explicitly targeting reliability on long-running, multi-step agent tasks — the failure mode that most often forces developers to babysit coding agents.
What Cursor says changed
In Cursor’s launch post, the company describes Composer 2.5 as its “most powerful model yet,” with improvements in intelligence and instruction-following for complex work.\
Key claims from Cursor’s announcement:
- Composer 2.5 is “more intelligent.”
- Composer 2.5 is “better at sustained work on long-running tasks.”
- Composer 2.5 is “more reliable at following complex instructions.”
Pricing (Standard vs Fast)
Cursor’s blog post includes explicit token pricing for two modes:
| Mode | Input tokens | Output tokens | Notes |
|---|---|---|---|
| Standard | $0.50 / 1M | $2.50 / 1M | Base pricing described for Composer 2.5 |
| Fast (default) | $3.00 / 1M | $15.00 / 1M | Cursor describes this as a faster variant with the same intelligence |
Launch promo: 2× included usage (this week)
Cursor says included usage for Composer 2.5 is doubled for the first week after launch.
Sources
- Cursor announcement post (May 18, 2026): https://x.com/cursor_ai/status/2056415413077233983
- Cursor blog post, “Introducing Composer 2.5” (published May 18, 2026): https://cursor.com/blog/composer-2-5
Read next
Keep building the workspace playbookCursor SDK Lands in Public Beta: Programmatic Agents in TypeScript with Local and Cloud Runtimes
Cursor launched the Cursor SDK in public beta on April 29, 2026, exposing the same agent runtime that powers Cursor desktop, CLI, and web behind a TypeScript package. `@cursor/sdk` lets you spawn agents against local files, Cursor-hosted VMs, or self-hosted workers, stream results, and bill on standard token-based pricing -- moving Cursor from an editor surface to a programmable platform.
Cursor vs Claude Code in 2026: Which AI Coding Tool Is Better?
Cursor is the better choice for developers who want AI integrated into a visual IDE with features like inline editing and background agents. Claude Code is the better choice for developers who prefer working in the terminal and need deep multi-file reasoning, shell integration, and agentic workflows.
Frequently Asked Questions
What is Composer 2.5?
Composer 2.5 is Cursor’s newest Composer model. Cursor says it’s a substantial improvement in intelligence and behavior over Composer 2, and that it’s better at sustained work on long-running tasks and more reliable at following complex instructions.
Is Composer 2.5 available now?
Yes. Cursor says Composer 2.5 is now available in Cursor.
What are the Composer 2.5 token prices?
Cursor’s blog post lists Composer 2.5 (Standard) at $0.50 per million input tokens and $2.50 per million output tokens, and a Fast variant at $3.00 per million input tokens and $15.00 per million output tokens.
Is Cursor offering any launch promotion for Composer 2.5?
Yes. Cursor’s announcement post says that included usage is doubled for the next week, and Cursor’s blog post similarly says Composer 2.5 includes double usage for the first week.