ChatGPT settings guide
ChatGPT (OpenAI) · General purpose · Multimodal
- Token burn
- Medium burn
- Paid plans from
- $20/mo (Plus)
Moderate; a few settings changes save money. Last reviewed September 2026; plans and model names change often, so check the tool's own pricing page before you buy.
Fill in Custom Instructions (Settings → Personalization) with who you are and how you want answers. It applies to every chat, so you stop repeating it.
Use the fast default model for everyday writing and questions, and switch to a Thinking model only for multi-step reasoning, analysis or code that must be right.
Keep related work in a Project so files and instructions stay together across chats.
Turn Memory on if you want ChatGPT to remember preferences across chats; review and delete saved memories in Settings from time to time.
For long code or document work, start a fresh chat per task. Very long threads carry old context that can lead the model astray.
Ask for the length and format you need ("under 120 words", "a table"). It is the simplest way to cut filler.
Use canvas for editing long documents or code in place instead of asking for the whole text to be rewritten each time.
API cost estimate
Monthly cost at list prices (September 2026), before and after a token cut you choose.
- Now
- $4.80
- ₹403
- After a 25% cut
- $3.60
- ₹302
- Saved per month
- $1.20
- ₹101
Assumes about 1,500 input and 500 output tokens per request (medium burn). Excludes caching discounts, search or request fees and taxes. Check the vendor's pricing page before budgeting.
Reveaxa guide reference: RVX-T-chatgpt
More Chat / LLM guides
Claude
Chat / LLM · Long context · Reasoning · Coding
Put standing context in a Project: upload reference files and write project instructions once, so you stop pasting the same background into every chat.
Gemini
Chat / LLM · Multimodal · Google Workspace
The free plan includes Gemini 3.6 Flash, with Gemini 3.1 Pro when capacity allows. Google AI Plus is ₹399/month and AI Pro ₹1,950/month in India (gemini.google, September 2026).
Groq / Llama 3
Chat / LLM · Ultra-fast · Free LLM
Groq serves open models very fast on its own LPU hardware; use it when response speed matters (real-time apps, prototyping, streaming).