ChatGPT vs Gemini: an honest comparison
ChatGPT is the most polished all-round assistant with the largest ecosystem; Gemini is the better bundle, with a 1M-token context window on paid plans and the lowest serious price floor at $4.99. Both are excellent and both have the same ceiling: you get one vendor's models, one linear chat shape, and a context the app assembles for you invisibly. We build NodePad, so read the third column with that in mind. It exists for the question these two never answer: what if you did not have to pick, and could hand either model exactly the context you chose? If you live in one ecosystem and chat casually, pick that ecosystem. If your AI conversations are work, the third column is the honest pitch.
| dimension | ChatGPT | Gemini | NodePad |
|---|---|---|---|
| Price of entry | Free; Go $8/mo; Plus $20/mo; Pro $200/mo. Ads are being tested on Free and Go. | Free; AI Plus $4.99/mo; AI Pro ~$20/mo; AI Ultra $100 to $200/mo, cut at I/O 2026. | Free on community models; paid plans from $5/mo. No ads on any tier. |
| Models you can use | OpenAI models only (GPT-5.5 family free, GPT-5.6 family paid). | Google models only, picked by capability label rather than exact version. | Claude, GPT, and Gemini in the same workspace; you pick the model per message. |
| Context window | Tiered and far below the API's: roughly 32K tokens on Plus, 128K on Pro; exact figures are not published. | Tier-gated: 32K free, 128K on Plus, 1M on Pro and Ultra; the biggest paid window of the big apps. | The model's full window, spent on context you chose rather than a resent scrollback. |
| Who assembles the context | The app: hidden system prompt, memory synthesis (rewritten in the background), then your truncated history. | The app: personalization from chat history and, if you opt in, your Google apps. | You. The model receives the path you built plus what you referenced or merged. Nothing hidden, nothing injected. |
| Branching a conversation | "Branch in new chat" copies context into a separate flat chat; no tree view, no way back. | None; conversations are strictly linear. | Fork from any point onto a visible parallel path; the original chat continues untouched next to it. |
| Working across chats | Projects group chats around shared files; the chats still cannot see each other. | No working folders or projects shipped as of mid-2026; the sidebar list is the organization. | Multiple chats live on one canvas; merge pulls several into scope for one answer. |
| Memory | The deepest memory system anywhere, spanning everything you have said; also the least inspectable. | Chat-history personalization with an on/off toggle; deletable activity items. | None by design. Continuity comes from the canvas you can see, not a profile you cannot. |
| Voice, mobile, media | Best-in-class apps, full-duplex voice, native image generation. | Gemini Live voice free, image and video generation, deep Android and Workspace integration. | Browser only; no voice, no image generation. If those are daily drivers, keep one of these apps alongside. |
| Usage limits | Dynamic and unpublished; they flex with load and region. | Published as multipliers only (2x, 4x, 5x, 20x), never absolute numbers. | A visible usage meter; when it runs out you add usage or wait for the reset. No hidden throttles. |
No credit card to start. Every paid plan reaches every model.