
Convert Claude Code and Codex Usage to API Dollars — Then Pick $20, $100, or $200
Run two local ccusage commands, read monthly API-equivalent Cost (USD), and map it onto current Claude Max and ChatGPT Pro seats.
Desk 03 of 08
Coding assistants, agent systems, harnesses, developer seats, and usage decisions.
I / Start here

Run two local ccusage commands, read monthly API-equivalent Cost (USD), and map it onto current Claude Max and ChatGPT Pro seats.
GitHub will start reopening credit-card and PayPal signups for Copilot Business and Enterprise on September 1. Existing customers on those payment methods shift to upfront seat billing from October 1, at their next billing-cycle start.
Copilot's included AI credits reset at 00:00 UTC on the first day of each calendar month. Seat billing can follow another date, creating a short first credit window for mid-month starts.
II / Complete desk
No guides match this format. Show all in this desk
GitHub Copilot credits and Claude Code weekly limits step down within one week. Use this 30-minute checklist to keep, cap, or switch before the deadlines.
Cursor now documents list-price billing for Auto, but its new Router modes are not yet on individual plans. A framework for pinning models, watching usage, or leaving.
Cursor documents different rules for individual refunds and Teams seat removal. Used Teams seats remain occupied through cycle end, while applicable adjustments appear as future invoice credit.
ChatGPT-login Codex loses GPT-5.4 and GPT-5.4 mini on August 31, while API access remains. Decide whether to migrate, switch, or remove the pin.
Notion Workers start using credits October 15, 2026. Measure Workers and Custom Agents, then clear paid seats before renewal.
A decision framework for burst-week freelancers after ChatGPT Plus restored a five-hour Codex and Work limit: stay on $20, reshape the sprint, or move to Pro.
Core monthly credits do not roll over, purchased packs expire after six months, and a paused Replit subscription can still bill Agent and hosting usage. Track each clock separately.
On legacy credit-based Windsurf Enterprise contracts, a replacement inherits the seat's used prompt credits. Current self-serve quotas and newer Enterprise ACUs use different meters.
As of August 2026, prompt skill is baseline. A freelancer can build a production-shaped agent with free or self-hosted infrastructure components; model inference is a separate cost or local choice.
A field test of running Grok, Claude, and Codex as one content pipeline on DeepSeek Harness—roles, real commands, subscription-quota cost, and when the extra steps are worth it.
A practical test for deciding whether agentic coding tools earn their keep on personal, unpaid projects—and where they quietly cost you.
A practical guide to choosing between autocomplete, chat-based, and agentic coding assistants based on your work, risk tolerance, and budget.
III / Continue browsing
Changed terms and deadlines worth checking. No promo blasts.
Subscriptions are not open yet. New guides are available in the library.
Browse guides