AI subscription limits compared
Message caps, context windows, file sizes and output limits across every major AI subscription — the numbers people search for, in one table. Approximate and change often; verify officially.
Message and context limits
| Subscription | Standard usage limit (approx.) | Context window | Max output (approx.) |
|---|---|---|---|
| ChatGPT | ~80 msgs / 3h | 128K tokens | ~16K tokens |
| Claude | ~45 msgs / 5h | 200K tokens | ~32–64K tokens |
| Gemini | generous daily cap | up to 1M tokens | ~8–64K tokens |
| Perplexity | 300+ Pro searches / day | ~128K tokens | ~4–8K tokens |
| Grok | high daily cap | 131K tokens | ~16K tokens |
| Copilot | priority, fair-use | 128K tokens | ~16K tokens |
| MultipleChat | per-model, generous | up to 200K tokens | per model |
File upload & type limits
| Subscription | Max file upload (approx.) | Files per chat (approx.) | Supported file types |
|---|---|---|---|
| ChatGPT | ~512 MB | ~10–20 | PDF, Office, CSV, images, code |
| Claude | ~30 MB | ~20 | PDF, text, CSV, images |
| Gemini | ~100 MB | varies | PDF, Office, Drive, images |
| Perplexity | ~50 MB | many | PDF, CSV, images, docs |
| Grok | ~20 MB | few | images, some docs |
| Copilot | ~30 MB | varies | Office, PDF, images |
| MultipleChat | up to 200 MB (higher tiers) | varies | All file types (PDF, Office, CSV, images, code + Mistral OCR) |
Images, web & languages
| Subscription | Image generation limit | Live web access | Languages |
|---|---|---|---|
| ChatGPT | daily cap | yes | 95+ |
| Claude | — | limited | 95+ |
| Gemini | daily cap | yes (Google) | 100+ |
| Perplexity | limited (Pro) | yes (core) | many |
| Grok | daily cap | yes (X + web) | many |
| Copilot | daily boosts | yes (Bing) | 95+ |
| MultipleChat | 8 models, per-plan | yes (Perplexity) | many |
What happens when I hit the limit?
On most plans you are not cut off — you are downgraded or paused. Hitting the cap on the flagship model usually means you keep working on a smaller/faster model, or you wait for the rolling window (often a few hours) to reset. Heavier tiers raise or remove these caps. Always confirm the current numbers on the official page before relying on them.
Official pricing & limits pages
Always verify current numbers here — they change frequently.
- ChatGPT: OpenAI pricing
- Claude: Anthropic pricing
- Gemini: Google Gemini plans
- Perplexity: Perplexity Pro
- Grok: xAI Grok
- Copilot: Microsoft Copilot
Last checked 14 June 2026.
AI limits — questions
What happens when I hit an AI usage limit?
On most plans you are not blocked — you are moved to a smaller/faster model or asked to wait for a rolling window (often a few hours) to reset. Higher tiers raise or remove the caps.
Which AI has the biggest context window?
Gemini offers the largest (up to ~1M tokens); Claude has 200K; ChatGPT, Copilot and Perplexity are around 128K. Bigger windows handle longer documents in one pass.
Which AI accepts the largest files?
ChatGPT accepts very large files (~512 MB); MultipleChat up to ~200 MB on higher tiers; most others are 30–100 MB. See the file upload limits page.