AI custom instructions character limits (2026 cheat sheet).
By Bernard Huang · Updated
Exact field lengths, byte caps, and file paths across ChatGPT, OpenAI Codex, Claude, Claude Code, Meta Muse, Gemini, Cursor, and Grok.
On this page
2026 custom instructions limit & path reference table.
| Platform & surface | Exact menu path or file location | Documented cap | Recommended tuning size |
|---|---|---|---|
| ChatGPT Custom Instructions | Settings → Personalization → Custom Instructions | 1,500 characters per longer field | 600–1,200 chars (use our compact generator) |
| ChatGPT Projects | Project → Project Instructions | Multi-kilobyte Markdown (overrides account custom instructions in that project) | 800–2,000 chars |
| OpenAI Codex & ChatGPT Work | ~/.codex/AGENTS.md + repo AGENTS.md / AGENTS.override.md | 32 KiB (32,768 bytes) combined across all files (project_doc_max_bytes) | Keep personal ~/.codex/AGENTS.md under 2 KiB |
| Claude Personal Preferences | Settings → General → Personal Preferences | Multi-paragraph text field (no 1,500-char lock) | 500–1,200 chars (6–10 bullets) |
| Claude Projects | Project → Project Instructions | Shared with project knowledge context window | 800–2,500 chars |
| Claude Code | ~/.claude/CLAUDE.md & ~/.claude/output-styles/*.md | Local Markdown file loaded into session context | 600–1,800 chars with keep-coding-instructions: true |
| Meta Muse | Settings → Personalization → Assistant Identity → Soul.md | Editable Markdown file (alongside Identity.md and Memory.md) | 500–1,500 chars inside How to work with me |
| Google Gemini | Settings → Saved info, Gems, or ~/.gemini/GEMINI.md | Short individual memories in Saved info; full Markdown in Gems and GEMINI.md | 1–2 sentences per Saved info item; 600–1,500 chars in Gems |
| xAI Grok & Grok Bot | Account response settings or Grok Bot Description | Compact text box in Bot Description; full system prompt in API | 500–1,200 chars |
| Cursor & GitHub Copilot | .cursor/rules/*.mdc & .github/copilot-instructions.md | Local repository Markdown files | 600–1,800 chars |
Why shorter tunings beat 10,000-word prompt walls.
Even when an app accepts large Markdown files, pasting thousands of words into global instructions hurts reply quality for three reasons:
- Instruction dilution: Every extra rule competes for attention with your live task prompt.
- Over-literal compliance: Newer reasoning models (like GPT-6 Astra and Claude Opus 5.5) follow system and
AGENTS.mdrules much more literally than older models did. All-caps“NEVER/ALWAYS” walls often cause agents to stop early or force a rigid format onto the wrong task. - Shared monorepo byte caps: In OpenAI Codex, the 32 KiB
project_doc_max_bytesbudget is shared across~/.codex/AGENTS.md, the repo rootAGENTS.md, and nested folderAGENTS.mdfiles. Keeping your personal file under 1 KiB leaves 97% of the budget for repo build and test commands.
A universal 580-character tuning block that fits every platform.
This 580-character block fits comfortably inside ChatGPT’s 1,500-character box and takes less than 2% of Codex’s 32 KiB budget:
Want a version matched to your own personality profile? Open the Custom Instructions Generator or pick any of the 43 profiles in the library and select ChatGPT · custom instructions for a ready-to-paste compact block.
Questions people ask.
What is the character limit for ChatGPT Custom Instructions in 2026?
OpenAI documents a 1,500-character limit for the longer Custom Instructions fields under Settings, Personalization, Custom Instructions. Every profile in AgentTune includes a reviewed compact export under 1,500 characters.
What is the byte limit for OpenAI Codex AGENTS.md?
Codex reads up to 32 KiB (32,768 bytes) combined across all AGENTS.md and AGENTS.override.md files in the directory chain, configured by project_doc_max_bytes.
Do ChatGPT Project Instructions have the same 1,500-character limit?
No. Project Instructions inside a ChatGPT Project accept full multi-paragraph Markdown tunings and override your global Custom Instructions inside that project.
How many rules should I put in Claude Personal Preferences or Meta Muse Soul.md?
Six to ten clear bullet points (roughly 500 to 1,200 characters) cover tone, answer order, formatting, and pushback without crowding out your task context.
What changed.
- First published with 2026 character and byte limits across ChatGPT, Codex, Claude, Muse, Gemini, Cursor, and Grok.