Cursor Launches Composer 2 with Frontier Coding Intelligence at Low Cost

Cursor

Mar 20, 2026 · Updated Apr 25, 2026

Cursor released Composer 2, its latest coding model, now available in Cursor. It scores 61.3 on CursorBench — up from 44.2 on Composer 1.5 — and is priced at $0.50/M input and $2.50/M output tokens.

Cursor released Composer 2, its proprietary coding model now available inside Cursor. Trained through continued pretraining and reinforcement learning on long-horizon coding tasks, it scores 61.3 on CursorBench and 73.7 on SWE-bench Multilingual — up from 44.2 and 65.9 for Composer 1.5. It can handle tasks requiring hundreds of sequential actions. Pricing starts at $0.50/M input and $2.50/M output tokens, with a faster variant at $1.50/M input and $7.50/M output as the default.

At ~60% on CursorBench with a median cost near $0.50 per task, Composer 2 undercuts GPT-5.4 — which scores in the 63% range at roughly $1.50 per task at its highest tier. Cursor describes this as a new optimal combination of intelligence and cost.

Try Composer 2 inside Cursor today or in the early alpha of the new Glass interface. Individual plan users access it through a standalone usage pool with generous usage included.

View the full update on cursor.com

Cursor

@cursor_aiMar 19

Composer 2 is now available in Cursor. https://t.co/Dl0EcpiLpf

812

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from Cursor →

Keep reading

Cursor Releases Composer 2 Technical Report on Coding Agent Training

Cursor published a technical report on Composer 2, a coding agent trained via pretraining on Kimi K2.5 and RL on real engineering tasks. It scores 61.3 on CursorBench — 37% above Composer 1.5 — matching frontier models at lower cost.

Cursor Publishes CursorBench, Its Internal Agentic Coding Evaluation Methodology

OpenAIMar 15

Cursor Publishes CursorBench, Its Internal Agentic Coding Evaluation Methodology

Cursor published CursorBench, its internal eval suite that scores models on real coding agent tasks from actual developer sessions. Public benchmarks struggle to differentiate frontier models reliably — CursorBench produces more separation where it matters most.

Cursor and Fireworks AI Detail the Specialized Training Infrastructure Behind Composer 2.5

Fireworks AIMay 27

Cursor and Fireworks AI Detail the Specialized Training Infrastructure Behind Composer 2.5

Cursor and Fireworks AI shared a technical breakdown of the distributed reinforcement learning infrastructure used to build the Composer 2.5 coding model. The team treats model weights as finite storage bits dedicated entirely to software engineering, allowing the model to match frontier performance at one-tenth the cost. This shift demonstrates how specialized products can use real-world usage as a proprietary training loop.

xAIJun 1

xAI Adds Cursor's Composer 2.5 to Grok Build for Long-Running Agentic Tasks

xAI announced that Cursor's Composer 2.5 model is now available inside Grok Build. The integration brings a specialized coding model to X Premium+ and SuperGrok subscribers, optimized for the sustained reasoning that long-running autonomous engineering tasks demand.