Got nerfed by Windsurf?
No receipts yet. Below: open-source replacements you can switch to today + tracked paid competitors ranked by who’s least likely to nerf you next. Step-by-step guides walk you through the actual install + config + API-key transfer - that part’s a Personal-tier feature.
Self-hosted, free, or freemium tools that cover the same workflow. Every entry is real software with a public repo and an active community.
Open-source Claude Code / Cursor replacement that lives in the terminal. Pick when you want zero vendor lock-in and any LLM backend.
- Terminal-native TUI coding agent (vim-like)
- Build/Plan mode toggle, session management
- Provider-agnostic: Claude/OpenAI/Google/local
curl -fsSL https://opencode.ai/install | bash (or npm i -g opencode-ai@latest)Permissive (MIT) open-weight model targeting ChatGPT/Claude-class chat and agentic coding without API lock-in - the strongest open option for a self-hosted stack.
- Fully-open (MIT) flagship MoE from Z.ai / Zhipu AI (weights released Jun 2026)
- ~744B total / ~40B active parameters, 1M-token context (~5x GLM-5.1)
- Strongest open-weight coding model at release - #2 on Code Arena, ~#3 on FrontierSWE behind only Fable 5 / Opus 4.8
ollama run glm-5.2 (or download weights from HuggingFace: zai-org/GLM-5.2, serve via vLLM/SGLang)ChatGPT's Code Interpreter, but local and against any model.
- Code Interpreter for any LLM, locally
- File ops + shell + browser control
- Python-first but multi-language
pip install open-interpreterNear-frontier open-weight model to back a self-hosted ChatGPT/Claude replacement or an agentic coding stack.
- Open-weight multimodal MoE from Moonshot AI: 1T total / 32B active
- 256K context, handles text + image + video
- INT4-native weights for cheaper serving
ollama run kimi-k2.6 (or download weights from HuggingFace: moonshotai/Kimi-K2.6, serve via vLLM/SGLang)A capable open-weight model for personal, self-hosted, or research chat/coding setups - if your use is non-commercial.
- 229B open-weight model from MiniMax
- Built for coding + agentic workflows (tool calling, agent teams)
- Served via vLLM/SGLang/Transformers
ollama run minimax-m2.7 (or download weights from HuggingFace: MiniMaxAI/MiniMax-M2.7, serve via vLLM/SGLang)Open-source Devin. Best when you want autonomous multi-step coding agents you control.
- Sandboxed agent that runs full dev tasks
- Web UI + CLI + headless modes
- Plug any LLM backend
docker run docker.all-hands.dev/all-hands-ai/openhands:0.30Other paid tools we track in the same category, ranked by current Nerf Index. Lowest = least likely to nerf you next.
1 critical receipt logged.
1 critical receipt logged.
1 major change on record.
Windsurf → Kimi K2.6 (open weights)
Unlock the full Kimi K2.6 (open weights) migration
Step one’s on us. The rest — config port, smoke test, the buried cancel path with a prefilled refund email, and the rollback — open the second you join.
- All 6 steps, copy-paste ready
- Every migration playbook — including ones we haven’t written yet
- Refund email that’s clawed back $80+
- gotnerfed ask — grounded help through every step