Free
$0/mo
Basic Cascade features
The agentic IDE from Cognition (the team behind Devin). Built on a VS Code fork around the Cascade agent, with proprietary SWE-1.5 models, SWE-grep / Fast Context retrieval, Codemaps for visual navigation, and 40+ IDE plugins. Multi-model access to Claude Sonnet 4.6, GPT-5.4, Gemini 3.1 Pro, and Windsurf's own SWE family.
+Pros
- Cascade agent with deep contextual awareness across the full codebase — not just open files
- SWE-1 proprietary model family optimized specifically for software engineering tasks
- Tight terminal and browser integration for true end-to-end task execution
- More affordable than Cursor at comparable feature tiers
- Cascade Flows enable multi-step automated workflows without manual intervention
- Multi-model support — Claude, GPT, Gemini, and the in-house SWE-1 family
−Cons
- Smaller community and extension ecosystem compared to Cursor and VS Code
- Not open source — fully proprietary IDE with no public codebase
- SWE-1 model still catching up to frontier models on the most complex reasoning tasks
- Fewer third-party integrations and MCP server support than Cursor
- Background agents less mature than Cursor's cloud VM-based implementation
Pricing
Pro
$20/mo
Premium models, more usage
Teams
$40/user/mo
Team management, SSO
Enterprise
From $200/mo
SOC 2, RBAC, BYOK, hybrid
Windsurf is the agentic IDE originally built by Codeium and acquired by Cognition — the team behind the Devin autonomous agent — in a roughly $250M deal in late 2025. The combined company is positioning Windsurf as the developer-facing surface while Devin runs in the background, with both products converging toward a single autonomous engineering platform.
Windsurf is built on a VS Code fork and reorganizes the editor around the Cascade agent, which plans multi-step edits, calls tools, and runs terminal commands with deep repository context. Under Cognition, Windsurf added proprietary models — SWE-1.5 reportedly runs ~13× faster than Sonnet 4.5 at near-frontier coding quality — plus SWE-grep / Fast Context, a retrieval system that surfaces relevant code up to 20× faster than embedding-based search, and Codemaps for visual codebase navigation. Premium models include Claude Sonnet 4.6, GPT-5.4, and Gemini 3.1 Pro.
Key Features
• Cascade: agentic flow that merges chat, planning, and multi-file editing
• SWE-1 / 1.5 / mini: proprietary models tuned for speed and codebase reasoning
• Fast Context (SWE-grep): non-embedding retrieval, up to 20× faster
• Codemaps: AI-generated visual maps of code structure and execution paths
• 40+ IDE plugins: JetBrains, Vim, Neovim, Xcode, and more (Tab + Chat)
• Multi-model: SWE-1.5, Claude Sonnet 4.6, GPT-5.4, Gemini 3.1 Pro
• Memories: long-running learning of project conventions
• Enterprise: SOC 2, SSO, RBAC, BYOK, hybrid deployment
Version History
SWE-1 model family introduced (SWE-1, SWE-1-lite, SWE-1-mini) — proprietary models optimized for software engineering. Cascade Agent improvements with enhanced multi-file awareness and deeper terminal integration.
Cascade Flows for multi-step automated workflows. Deep terminal integration and multi-repo context support. Improved browser integration for end-to-end task execution.
Initial release by Codeium as a standalone AI IDE. Introduced Cascade agent with full-codebase context, Supercomplete inline suggestions, and deep editor integration.
Signature Snippet
Open a Python repo in the Windsurf Editor, ask Cascade: "Add a /webhooks endpoint, update the router, write tests, and document it." Cascade plans the change, edits 5 files in one pass, runs the test suite, and reports back with a diff and the failing-then-passing test output.More in this Space
Prime Agent (Prime Intellect)
Prime Agent is an open-source (MIT) self-improving coding and research harness by Prime Intellect, built on a Recursive Language Model that runs inside a persistent IPython REPL and a Continual Harness that refines its own state via /refine. It took the same frontier weights from 30.2% to 95.5% on ARC-AGI-3 RHAE Best@1, powered daemon-backed sessions, and ships no hosted offering.
OpenCodeReview (ocr)
OpenCodeReview (ocr) is Alibaba's open-source AI code-review CLI. It reads git diffs and drives an LLM agent through a hybrid architecture where deterministic pipelines handle file selection, rule matching, and comment positioning while the agent does the reasoning. Built on two years of internal use at Alibaba scale, it ships a multi-language ruleset and consumes roughly 1/9 of the tokens of general-purpose agents.
Delta (by Zed)
Multiplayer environment from Zed for coding with agents. Threads unify the agent conversation, the code edits, and human review in one shareable artifact, replacing the branch/PR loop. Built on DeltaDB, a CRDT-based version control layer that records every edit between commits alongside the prompts and reasoning that produced them.