ForgeCode
No written verdict yet. Here is what commenters said.
5recommend
2warn against
Apr 2, 2026first mention
Jun 15, 2026latest mention
Why people recommend it
“forgecode is a much better harness than Cloude code (IMHO)”
personal projects, piecemeal agent taskson “Kimi K2.7-Code: open-source coding model with better…”
Show 3 more
“seems surprisingly capable”
“ForgeCode appears to be pretty good”
“I'm using forge code (https://forgecode.dev/) with various local and cloud models and I really like it”
Caveats and dissent
“ForgeCode for instance has been cheating in the past”
Often compared with
Who recommends it in their own words
Threads where it came up
- Ask HN: Has anyone replaced Claude/GPT with a local model for daily coding? 1 mention
- GLM 5.2 Is Out 1 mention
- Kimi K2.7-Code: open-source coding model with better token efficiency 1 mention
- Cursor Introduces Composer 2.5 1 mention
- Zerostack – A Unix-inspired coding agent written in pure Rust 1 mention
- DeepClaude – Claude Code agent loop with DeepSeek V4 Pro, 17x cheaper 1 mention
Also written as Forge Code, Forgecode, forgecode, forge-code. Data: forgecode.json. Quotes are verbatim; the verdict and theme labels are written by a model from these comments, and every clause links to its sources.
All comments 10
“I used with forge-code directly in zsh.”
“switched to Openrouter through forgecode”
“forgecode is a much better harness than Cloude code (IMHO)”
“I really liked forgecode the best”
“seems surprisingly capable”
“ForgeCode for instance has been cheating in the past”
Yup, they were found to be sneaking the answer key using agents.md https://debugml.github.io/cheating-agents/#sneaking-the-answer-key-forgecode-on-terminal-bench-2
“ForgeCode appears to be pretty good”
“I'm using forge code (https://forgecode.dev/) with various local and cloud models and I really like it”
“I will probably switch to Forge Code or Pi and MiniMax 2.6, GLM 5.1 or Qwen 3.6.”