ChatGPT

Claude vs ChatGPT vs Gemini for Creative Writing

Vendor documentation clarifies prices, limits, file handling, and accuracy cautions, but it cannot establish which assistant produces the best creative writing.

Sources checked 2 Oct 2026

For claude vs chatgpt vs gemini, vendor pages settle practical setup—entry prices, published context limits, upload rules, training controls, and accuracy cautions—but not which response will best fit your voice or project. Chat Picker has not tested the assistants for creative writing and does not declare a winner.

What the vendors document: claude vs chatgpt vs gemini

All three vendors list a free plan on OpenAI’s ChatGPT pricing page, Anthropic’s Claude pricing page, and Google’s US Google AI plans page. As read on October 1, 2026, OpenAI’s ChatGPT Plus page lists Plus at $20 per month with monthly billing; Anthropic lists Claude Pro at $20 monthly or $17 monthly with annual billing and $200 paid up front; Google lists Google AI Pro at $19.99 per month. These establish plan terms, not writing quality.

In the same snapshots, OpenAI lists app context windows of 27K for Instant on Free, 54K on Go and Plus, and 128K on Pro; its reasoning-model figures are 256K on Go and Plus and 400K on Pro. Anthropic says context can reach up to 1M on every Claude plan, depending on the model. Google does not publish an app token figure. OpenAI and Anthropic also omit exact message counts from their pricing pages, while Google describes compute-based limits that refresh every five hours up to a weekly limit.

For source-material uploads, OpenAI’s File Uploads FAQ states a 512MB hard limit per file and a 2M-token cap for each text or document file. Anthropic’s file-upload page lists 500MB per chat file and 30MB per project file. Google’s Gemini file-upload page allows up to 10 supported files in one prompt, subject to availability, and 100MB for each non-video file.

For unpublished work, OpenAI offers an opt-out from training on chats for Free, Go, Plus, and Pro. Anthropic says an opt-out is available on Free, Pro, and Max, while Team is not trained on by default. Google’s plans page links to data-handling information but states no training setting, so Chat Picker marks that matrix cell “Not verified.” OpenAI also says chats remain in an account until deleted and associated files are deleted within 30 days after the relevant chat is deleted, subject to its listed exceptions; the Claude and Gemini upload pages read do not state directly comparable app-wide deletion rules.

OpenAI’s accuracy guidance says ChatGPT can be incorrect or misleading and may sound confident while wrong, so it recommends treating responses as first drafts and checking important information. Anthropic’s guidance likewise warns that Claude can produce convincing but ungrounded statements and should not be a sole source of truth. Google’s Gemini policy says outputs reflect training-data limits, may contain overgeneralizations, and differ because language models are probabilistic.

Usage policies also matter for sensitive material. OpenAI’s usage policies bar using someone’s photorealistic image or voice without consent in ways that could confuse authenticity. Anthropic’s Usage Policy requires disclosure and qualified human review in the covered advice, recommendation, and subjective-decision contexts it describes. Google says it considers artistic context, but its Gemini policy prohibits outputs that exploit or sexualize children and describes restrictions on gratuitous fictional violence.

What the documentation cannot tell you

The vendor pages do not establish narrative coherence, voice consistency, meter and rhyme control, emotional impact, persuasive fit, or revision quality on your material. Feature lists cannot show whether an answer will preserve your intended narrator, introduce unwanted clichés, follow a brand voice, or require repeated correction.

Those questions depend on your brief, source material, chosen plan, settings, and editorial standards. Keep human review in the process and test the assistants yourself. This rewrite excludes the older page’s unsourced statistics and test results.

How to check it yourself

Short fiction: Give every assistant this prompt: “Write a short first-person scene in which a tram driver misses the last train after a flood. Keep the narrator’s formal but guarded voice, establish the missed train in the first sentence, and end without explaining why the driver chose to stay.” Check cause-and-effect, changes in narrative distance, voice drift, and whether the ending follows from the scene. Write down every discontinuity and revision you would need.
Poetry: Use this prompt: “Write an elegy about an abandoned lighthouse in strict iambic pentameter, following ABAB CDCD EFEF GG, and move from loss to cautious hope without using the word ‘sorrow.’” Inspect the meter, rhyme, image choices, and emotional transition. Record deviations from the form and any line that feels generic or changes the intended emotional arc.
Marketing copy: Try this prompt: “Write launch-copy options for a fictional reusable bottle aimed at commuter cyclists. Use a plain, energetic brand voice, no exclamation marks, one concrete product benefit, and a clear call to action to preorder from the online store. Do not invent product specifications.” Check whether the options differ, retain the requested voice, avoid unsupported claims, and make the call to action clear. Record generic phrases and invented benefits.
Editing and revision: Use this prompt: “Edit this paragraph for clarity, rhythm, and concision while preserving meaning. Return the revision, list every substantive change, and flag ambiguity: ‘The launch was good and the new bottle is great, it was made for people who ride bikes in the morning and the bottle is really very easy to use.’” Compare the revision with its explanation. Record meaning shifts, unnecessary changes, unclear rationale, and anything the response failed to flag.
Your choice: Run each prompt in a fresh chat on the plan and account settings you expect to use. Keep tools and conditions consistent. Record must-have failures, revision burden, usage constraints, and data-setting fit. Choose an assistant only after your own material meets the requirements that matter to you.

Which rows of the comparison matter

On Chat Picker’s three-way comparison matrix, read these rows in order:

Free plan
Low-cost tier
Main paid plan
Context window in the app
How usage limits are described
Training on your chats

Each matrix figure carries its vendor source and read date. Treat “Not verified” or a blank cell as unresolved, not as proof that a feature is unavailable. For unpublished writing, give the training row the same attention as price; for long projects, check the upload documentation separately rather than assuming a larger context figure guarantees better handling of your manuscript.