After 6 months of coding with Hermes, DeepSeek V4 Flash, and Cline, this user ranks the tools
SafeImprovement7204 · reddit · 2026-07-27
A Reddit user compares a long list of AI coding harnesses and IDEs after six months of daily work on medium-to-large projects, including Flutter mobile apps, full-stack Flutter + C+ React systems, and hobby builds like an OpenRouter clone.
Main takeaways
- Hermes + DeepSeek V4 Flash has become their default setup and performs better than most agnostic harnesses.
- The key difference is Aphrodite: default Hermes often stalls, while Hermes paired with Aphrodite “just works.”
- Their rough score for Hermes:
- code fixing: 5–7/10
- new features: 7/10
- random tasks: 8/10
Other tools mentioned
- OpenCode with Xiaomi MiMo V2.5 Pro: looping responses or dead stops.
- pi.dev: decent, but uses more tokens than Hermes for similar work.
- Cline: strong planning and Kanban flow, but they have not set it up fully yet.
- Factory.io / Cursor: good but too expensive in token burn.
- Antigravity: best value among paid tools tried.
- GLM 5.2: noticeably better code quality than MiMo V2.5, Kimi 2.7, or DeepSeek V4, but too costly.
- Kimi 2.7 Code: large quota, weak code quality; better for research and building skills.
- Blackbox: desktop/CLI loops and hallucinates, though the API works fine.
- GitHub Copilot: burns budget fastest and is rated poorly.
The user is now asking for a new tool that is specifically strong at coding, and also wonders whether super-agentic.ai is legitimate or just a rebrand.
Related event: Developer Ranks AI Coding Tools After Six-Month Test(2 posts)→
More from coding & agent
- Memoria open-sources Git-style version control for AI agent memory — bigaiguy · 2026-07-27
- Open-source jcode claims 14 ms startup and multi-agent repo coordination — thisguyknowsai · 2026-07-27
- New ComfyUI prompt manager node adds on-the-fly edits, randomization, and shared presets — 3deal · 2026-07-27
- VS Code Remote makes the Claude-and-Codex server setup usable anywhere — nico_jeannen · 2026-07-27
- Developer runs Claude and Codex on a server to work from a phone or iPad — nico_jeannen · 2026-07-27
- Tanuki Context claims up to 94% token savings by turning some LLM inputs into images — 0syna · 2026-07-27