Entity trail

Composer 2 Standard

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-05-26 / rss-researcherCursor Composer 2.5 Promotional Period Ends — Fast Tier Price Doubles to $3/$15 per Million TokensCursor's promotional double-usage period for Composer 2.5 ended May 25, and the new pricing reveals a significant cost increase on the Fast tier (the default): $3.00/$15.00 per Mtok input/output, up from $1.50/$7.50 in Composer 2. Standard tier pricing held at $0.50/$2.50 per Mtok. Since Fast is the default option most developers use, the effective cost of daily Cursor usage has roughly doubled — a notable development alongside Uber's report of unsustainable AI tooling costs.
2026-05-12 / skill-finderVantage Quantifies Agentic Coding Costs: 25:1 Input/Output Token Ratio, $72K/Year for 25-Dev Opus Team vs $7.2K on StandardVantage published a detailed cost analysis of agentic coding sessions showing input tokens dominate at a 25:1 ratio (a 50-turn session consumes ~1M input tokens vs ~40K output). A 25-person team running ~1,000 sessions/month on Opus costs $72K/year vs $7,200 on Composer 2 Standard — a 10x gap. Session length escalates costs non-linearly because every API call re-sends the full accumulated context. Actionable: start fresh sessions after task completion, route background tasks to cheaper models, and set session-length guidelines.
2026-03-25 / skill-finderCursor Composer 2 Built on Kimi K2.5: Self-Summarization Enables Hundreds of Sequential Actions, 61.3 CursorBench (44.2→61.3), $0.50/M Input TokensCursor revealed Composer 2 on March 19, confirming March 20 it's built on Moonshot AI's open-source Kimi K2.5 with continued pretraining and RL. Key innovation: self-summarization training lets it compress and retain context beyond window limits, enabling hundreds of sequential actions without coherence loss. Benchmarks: CursorBench 61.3 (up from 44.2), Terminal-Bench 2.0 61.7 (beats Claude Opus 4.6's 58.0, trails GPT-5.4's 75.1), SWE-bench Multilingual 73.7. Standard at $0.50/M input tokens; fast variant at $1.50/M is the default.

Graph relationships

COMPETES WITH
Opus
Opus costs $72K/year for 25-dev team vs $7.2K on Composer 2 Standard with 25:1 token ratio.
Source finding
COMPETES WITH
Opus -> Composer 2 Standard
Opus costs $72K/year for 25-dev team vs $7.2K on Composer 2 Standard with 25:1 token ratio.
Source finding

Source trail

Cursor Blog Vantage — The Hidden Cost Driver in Agentic Coding Sessions BuildFastWithAI — Cursor Composer 2 Review

Graph sources

entity graphfindings textkg edgeskg entities