AI Potluck
Product / UX / Orchestration & agents

Windsurf

Codeium

AI-native IDE, forked from VS Code, built around an agentic flow engine that keeps context across multi-step coding tasks and suggests and executes terminal commands, file edits and debugging steps. It runs multi-model and multi-agent over the Agent Client Protocol, blending inline suggestions with autonomous agent actions.

Formerly Codeium, then Windsurf; acquired by Cognition in 2025 and now shipping as Devin Desktop, with windsurf.com redirecting there. Verified 2026-08-13 via the Devin Desktop product page and Cognition's acquisition announcement.

Openness

1 high confidence
1.0
source
closed
license
proprietary(SaaS, Cognition)
self-host
no

Proprietary agentic IDE; now part of Cognition's Devin Desktop. Closed source, subscription SaaS.

  • https://devin.ai/desktop recorded 2026-08-13

    Windsurf shipped as Devin Desktop, a proprietary agentic IDE sold on Free/Pro/Teams/Enterprise tiers; no source and no self-host option

  • https://cognition.ai/blog/windsurf recorded 2026-08-13

    Cognition's own acquisition announcement - "the acquisition includes Windsurf's IP, product, trademark and brand"

Adoption

4 medium confidence
4.0

The Devin Desktop (Windsurf) page reports "1M+ Users" and "4000+ Enterprise Customers", which places it in the 1-10M band, and Cognition-reported Windsurf ARR of about $82M corroborates commercial scale. The user total is a vendor headline with no period attached, so it is not an active count and the level rests on reported traction rather than a counted active-user figure.

  • https://devin.ai/desktop recorded 2026-08-13

    stats block - "1M+ Users - trusted by over a million developers worldwide" and "4000+ Enterprise Customers"; no period attached to the user figure

  • https://cognition.ai/blog/windsurf recorded 2026-08-13

    Windsurf brought $82M of ARR and a fast-growing business with enterprise ARR doubling quarter-over-quarter

Capability

4 medium confidence
4.0

Mature agentic-IDE feature set (multi-agent orchestration, model-agnostic via ACP) but no first-party SWE-bench score published; scored on feature matrix, a notch below the benchmark anchor.

  • https://devin.ai/desktop recorded 2026-08-13

    Agent Command Center, multi-agent and multi-model work over the Agent Client Protocol, Fast Context codebase retrieval, Supercomplete; no benchmark figure published

Verified 2026-08-13