Sol 5.6 on $20 Codex did 2.2x more average tasks than Fable 5 on $20 ClaudeCode d devforth.io 3 deviscool 4 months ago 0 comments
Real limits converted to API-equivalent $ value for Claude Code, Codex, Copilot t twitter.com 1 deviscool 4 months ago 1 comments
Reverse-engineered token Value of coding agent Plans d devforth.io 1 deviscool 4 months ago 1 comments
Show HN: AdminForth – Open-Source Admin Framework for Node.js and Vue 2 deviscool 1 year ago 0 comments
Even open source libs now care about LLMs as a basic feature a adminforth.dev 1 deviscool 2 years ago 0 comments
Admin Panel in Minutes: AdminForth is Vue.js Framework for Rapid CRUD creation a adminforth.dev 1 deviscool 2 years ago 1 comments
Think about it as about much you would pay in API costs if the agent used the official API directly instead of being included in a monthly plan.
Insights:
* Personal Codex Plus ($20/mo) gives about $130 in monthly token value, while Business Codex ($25/mo) gives about $25
* Claude Code gives the same $1,200 token value on a Team Standard seat ($25/mo) as on the personal plan ($20/mo).
* After the June update, Copilot Pro ($10/mo) gives only about $15 in token value, so it no longer makes much sense.
* Google AI Pro per $20 gives about $171 in token value on Gemini models + $40 on several 3rd-party models