Windsurf

Codeium's AI-native IDE positioned against Cursor — in 2025, OpenAI reportedly acquired it for ~$3B, pushing the IDE war to the next tier.

1. Core Product / Service

Windsurf is the desktop IDE Codeium launched in 2024-11 (also forked from VS Code), positioned as "the first agentic IDE that flows with you". Codeium previously made its name on enterprise code-completion tools (a free / enterprise alternative to GitHub Copilot).

Headline features:

• Cascade: Windsurf's signature multi-step agent workflow that can read + write across files + run tests + fix bugs automatically, considered "smoother than Cursor Composer"
• Supercomplete: cross-file context completion (vs Cursor Tab)
• Codebase awareness: indexes the entire repo locally
• Memories: remembers project preferences / style across sessions
• Multi-model: Claude / GPT / Gemini selectable

Pricing (2026-05):

• Free — 50 prompt credits / month, limited to GPT-4o-mini / Cursor-small-class
• Pro $15/mo — 500 credits + advanced models
• Pro Ultimate $60/mo — 1500 credits + Cascade high allowance (after late-2025 adjustment)
• Teams $30/seat·mo — SSO + multi-user sharing
• Enterprise — negotiated

Pricing is slightly lower than cursor ($15 vs $20) — Codeium's long-running "price war" style.

2. Target Users & Pain Points

• Developers who want to try Cursor but don't want to pay $20: Windsurf's $15 entry is cheaper
• Enterprise: Codeium's Enterprise business is long-established, with a large Fortune 500 client base (including finance / defense / healthcare); they want compliance + on-prem deployment; the Windsurf IDE is a free upgrade for those customers
• VS Code users: Windsurf's UX is closer to original VS Code, with lower migration friction than Cursor

3. Competitive Landscape

Windsurf and Cursor are the two strongholds of the 2025 IDE war; Windsurf differentiates slightly on product polish and enterprise penetration (more enterprise-heavy), Cursor is more mainstream among indies / startups.

4. Unique Observations

• OpenAI acquisition rumors / partial confirmation (Spring - late 2025): The Information / Bloomberg have repeatedly reported that OpenAI and Codeium are in acquisition talks at ~$3B. The motivation behind it is that OpenAI wants to own its own IDE entry point (not handing it to Anthropic via Cursor, and not letting Microsoft Copilot dominate alone). As of 2026-05-10, the deal status is uncertain at this wiki point — some public reports say the deal has been signed but is blocked by Microsoft's OpenAI agreement clauses (IP rights), some say it's been completed. Keeping status: active, confirm in a future wiki audit.
• If the OpenAI acquisition holds: it means an IDE owned by chatgpt in addition to Cursor; github-copilot is in the Microsoft camp, Windsurf is in the OpenAI camp, Cursor is independent — a three-way standoff.
• Strategic meaning of the $15 price: Codeium has long used free / cheap individual developers as a customer acquisition tool, with the real money coming from Enterprise. Windsurf $15 / Pro is 25% cheaper than Cursor — designed to slow Cursor's individual-side growth.
• Token consumption / gross margin: Windsurf's credit system exposes cost more explicitly than Cursor — one Cascade prompt with Claude Sonnet 4.7 typically eats 10-50 credits, and heavy users burn through a $60 tier in a week. Same issue as Cursor with "model price hike → must adjust tiers"; Windsurf adjusted pricing during 2025 too.
• Distribution moat: relatively thin. Codeium's Enterprise channel is the real moat (1000+ enterprise customers); on the individual side, Windsurf IDE relies on "VS Code-like + cheap"; if not acquired by OpenAI, standalone Windsurf likely can't outrun Cursor long-term.
• Product philosophy vs Cursor: Cursor leans "the more aggressive the agent the better" (Composer multi-file edits); Windsurf Cascade leans "agent collaborating with you" (each step can be intervened / rolled back) — the style difference is a creed war among heavy users.
• In the acquisition scenario, what happens to Codeium parent: Codeium's original Enterprise business (selling code completion + self-hosted per seat) has substantial volume; industry chatter is that the two businesses may split — OpenAI takes the IDE part, the Enterprise business stays independent or is spun off.

5. Financials / Funding

Parent Codeium / Exafunction:

• Series A $18M (2022, Greenoaks lead)
• Series B $65M (2023-12, Kleiner Perkins lead)
• Series C $150M (2024-08, General Catalyst lead, valuation $1.25B)
• Series D (Spring 2025 rumored ~$150M, valuation in the $3B range)
• OpenAI acquisition (Late 2025, ~$3B, status pending latest reports)
• ARR: not fully public; industry estimate late-2025 at $80-150M (with Windsurf IDE share rising month over month)

6. People & Relationships

• Founders: Varun Mohan (CEO), Douglas Chen — MIT-trained engineers, originally worked on GPU inference optimization
• Investors: Greenoaks, Kleiner Perkins, General Catalyst
• Underlying models: claude-ai, chatgpt, gemini, Codeium's in-house completion model
• Competitors: cursor, github-copilot, Claude Code, replit
• Potential parent (2025-2026): OpenAI (acquisition pending final confirmation)
• Key relationships: if acquisition completes, will be under the OpenAI umbrella with chatgpt, potentially sharing underlying model cost structure

Sources

• https://codeium.com/windsurf (2026-05-10)
• https://windsurf.com/pricing (2026-05-10)
• https://www.theinformation.com/articles/openai-windsurf-acquisition (2026-05-10)
• https://www.bloomberg.com/news/articles/openai-codeium-deal (2026-05-10)