Cursor
AI-native IDE vendor; product integrates frontier models into developer workflows with local code context.
Cursor is not a foundation-model lab; its economic story is software: retention, seat expansion, and the marginal cost of inference inside an editing loop. Mid-2026 debates center on whether IDE copilots are a category or a feature—Microsoft, JetBrains, and open-core editors all compete for the same keystrokes.
We list the model families most central to Cursor’s shipped experience and track claims about developer productivity with an insistence on measured outcomes, not anecdotal speed-run videos.
Models shipped
- GPT-4 — *Multimodal generalist that reset 2023 expectations for instruction-following and tool use at scale.*
- Claude 3.5 Sonnet — *Anthropic mid-2024 workhorse tuned for coding and long-context tasks at competitive latency.*