### Paying Fix Rate Monthy?
If you're on a fixed rate, monthly payment plan, your AI provider will cut you off if you hit your token usage limit before the reset time.
Monitor your Claude Code usage here:
https://claude.ai/settings/usage
If you're using another platform, find that platform's usage page and keep it open in a browser at all times.
### Paying via API usage (even WORSE!):
If you're vibe coding with an API subscription, you're paying way WAY too much! Switch to a monthly plan and save yourself TENS OF THOUSANDS of dollars per month!
### Here's how to cut your token usage:
1. Switch models to a less expensive model for tasks that don't require heavy AI (ticket management, local file search, summarizing, etc..., basically everything except planning and writing more than boilerplate code and simplistic unit tests)
2. Have at least TWO AI sessions open; 1 for coding, one for ticket management. This ALONE will cut back on the context window length for every round trip.
3. Set the model for ticket management to the lower AI model.
4. If you have the hardware to run local AI, INSTALL OLLAMA AND A GOOD MODEL. Install ollama MCP tools and have a rule in your CLAUDE.md (or your AI's equivelent) that it MUST use the ollama mcp tool(s) when doing anything other than heavy cognitive load (Planning & complex coding).
5. Set a rule in CLAUDE.md that when possible, it should write script files for researching problems and have the script look at the responses of queries and such and make decisions based on the results, rather than having your expensive AI tokens manually doing everything.
6. Tell it when writing scripts that it can ADD AI capabilities in it via local Ollama.
---
*Note: There was no AI used in writing my article above. All human written. Below is a ChatGPT generated list of token usage pages for all the big AI platforms:*
---
* **Anthropic Claude**
* Web UI: [https://claude.ai/settings/usage](https://claude.ai/settings/usage)
* API: [https://console.anthropic.com/usage](https://console.anthropic.com/usage)
* **OpenAI ChatGPT**
* Web UI: [https://chatgpt.com/settings/usage](https://chatgpt.com/settings/usage)
* API: [https://platform.openai.com/usage](https://platform.openai.com/usage)
* **Google Gemini**
* Web UI: none
* API: [https://console.cloud.google.com/apis/api/generativelanguage.googleapis.com/metrics](https://console.cloud.google.com/apis/api/generativelanguage.googleapis.com/metrics)
* **Microsoft Copilot**
* Web UI (personal): none
* Admin / API: [https://admin.microsoft.com/Adminportal/Home#/reportsUsage/CopilotUsage](https://admin.microsoft.com/Adminportal/Home#/reportsUsage/CopilotUsage)
* **Groq (GroqCloud)**
* Web UI: none
* API: [https://console.groq.com](https://console.groq.com)
* **DeepSeek**
* Web UI: none
* API: [https://platform.deepseek.com/usage](https://platform.deepseek.com/usage)
* **Perplexity AI**
* Web UI: [https://www.perplexity.ai/settings](https://www.perplexity.ai/settings)
* API: none
* **xAI Grok**
* Web UI: none
* API: none
* **Meta AI**
* Web UI: none
* API: none
* **Mistral AI**
* Web UI: none
* API: [https://console.mistral.ai/usage](https://console.mistral.ai/usage)
* **Cohere**
* Web UI: none
* API: [https://dashboard.cohere.com/usage](https://dashboard.cohere.com/usage)
* **Together AI**
* Web UI: none
* API: [https://api.together.ai/settings/usage](https://api.together.ai/settings/usage)
* **Fireworks AI**
* Web UI: none
* API: [https://fireworks.ai/account/usage](https://fireworks.ai/account/usage)
* **Replicate**
* Web UI: none
* API: [https://replicate.com/account](https://replicate.com/account)
* **Hugging Face**
* Web UI: none
* API / Billing: [https://huggingface.co/settings/billing](https://huggingface.co/settings/billing)
* **OpenRouter**
* Web UI: none
* API: [https://openrouter.ai/activity](https://openrouter.ai/activity)
* **Poe (Quora)**
* Web UI: none
* API: none
About to Run Out of Claude Tokens? Do This NOW!
By Mike
173 views
0