Just released my final Python build for Doorman.
Moving forward the gateway will be re-written in Rust, controls in Python, UI in Next.js for v2 and on.
I originally wrote the gateway in Python for its ease of use and to practice my Python skills. Ignoring its limitations. More to come :)
https://t.co/V2hPpBPorb
Dropped a beta release of BetterCode! Combine your Claude, Codex, and Gemini CLIs into one! Local pre processing to pick the best model, run code reviews, and more!
https://t.co/T1FuxioKUn
Considering making this fully open source and charging enterprise $2/month/user for account management, analytics, and some other features.
Not sure what to do.. thoughts?
BetterCode is launching for FREE on April 8th!
Too many of us are selecting GPT 5.4 Extra High and Opus 4.6 Thinking for all of our prompts.
BetterCode runs 100% local to your machine and selects the best model for your task.
Use your existing Codex and Claude subscriptions!
But yes, there are some differences when using Claude compared to Codex or Gemini. The app is essentially an all in one workspace that acts as a terminal pass through. Local LLM pre processing -> Auto selected LLM for implementation (actual companies CLI is opened within BetterCode), then some local LLM post processing.
At the moment Claude is limited to be selected manually or using an API key (if using auto select). Since they don’t allow system orchestration.
Experimenting with making the user click “Run in Claude” after pre processing is complete and if a claude model is selected by BetterCode.
@PratikSinhatwt No reason to pick anymore.. You can use @bettercodeapp to pick for you. Its FREE. Behind the scenes it runs a small local model to pick the best model given the prompt. Use it with your Claude, Codex, and/or Gemini subscription or API key.
@thekitze Why be forced to pick? I'm building @bettercodeapp to run a small local model to decide for you. Save on tokens, context, and keep usage low. Works with your Claude, Codex, and Gemini subscriptions. For FREE.
BetterCode is launching for FREE on April 8th!
Too many of us are selecting GPT 5.4 Extra High and Opus 4.6 Thinking for all of our prompts.
BetterCode runs 100% local to your machine and selects the best model for your task.
Use your existing Codex and Claude subscriptions!
Everyone's having issues with AI usage limits. So I started building @bettercodeapp.
Use your Claude, Codex, and Gemini subscriptions or use an API key.
Pre/Post processing with a local model, auto model selection based on the given task, multi-agent code reviews, and more!
STOP RUNNING OUT OF AI USAGE!
BetterCode runs a local model to pre-process and determine the best model for the job.
Utilize your monthly Claude, ChatGPT, or Gemini subscrions or API token!
https://t.co/NHEfr5TQKR
STOP RUNNING OUT OF AI USAGE!
BetterCode runs a local model to pre-process and determine the best model for the job.
Utilize your monthly Claude, ChatGPT, or Gemini subscrions or API token!
https://t.co/NHEfr5TQKR