Claude vs ChatGPT vs Gemini: Multilingual Support

Vendor pages document language availability, plan limits, prices, and data controls, but not comparative multilingual quality; Chat Picker has not tested the assistants.

Sources checked 2 Oct 2026

If you are weighing Claude vs ChatGPT vs Gemini for multilingual work, the vendors’ own pages settle the surrounding terms: language settings or availability, input modes, prices, limits, and data controls. They do not provide a shared test of accuracy, cultural nuance, right-to-left handling, or speed, and Chat Picker has not tested the assistants for this use. Compare the documented terms, then test your own material.

What the vendors document for Claude vs ChatGPT vs Gemini

As read on 2026-10-01, Google’s Gemini availability page says the Gemini web app is available in more than 70 languages and over 230 countries and territories. OpenAI’s language-setting page says ChatGPT detects the browser or mobile device language for supported languages and lets logged-in users select a preferred language. Anthropic’s supported-countries page covers Claude.ai availability but gives no comparable language count. These are access facts, not comparative evidence about language nuance.

For spoken input, OpenAI says Voice availability depends on the plan, workspace, region, app version, and parental controls. Claude calls Voice a beta feature across plans and says conversations count toward regular usage limits. Gemini Live requires Android, is unavailable on the web and in Google Messages, and is not supported with Farsi in the Gemini mobile app.

The linked pricing pages, read on 2026-10-01, show:

ChatGPT pricing: Free includes unlimited text chats but limits uploads, images, Voice, and deep research. Go is A$13/month on the Australian page and may include ads; its US dollar price is not verified. Plus is $20/month.
Claude pricing: Free includes web search, file creation, code execution, and memory. No low-cost tier is listed between Free and Pro. Pro is $20/month, or $17/month with annual billing and $200 up front.
Google AI plans: Free requires a Google Account and lists 15 GB of storage. AI Plus is $4.99/month with 2x the Free usage limits and 400 GB; AI Pro is $19.99/month with 4x the Free limits and 5 TB.

Training settings also differ. ChatGPT offers an opt-out on Free, Go, Plus, and Pro. Claude offers an opt-out on Free, Pro, and Max, while Team is not trained on by default. The Gemini plans page links to Google’s data-handling information but states no setting, so Chat Picker marks that point “Not verified.”

On the same pages, read on 2026-10-01, ChatGPT lists 27K context for Free, 54K for Go and Plus, and 128K for Pro on instant models; its reasoning-model figures are 256K for Go and Plus and 400K for Pro. Claude says context can be up to 1M on every plan and varies by model. Gemini’s plans page gives no app token figure. ChatGPT and Claude publish no exact message counts there; Gemini uses compute-based limits that refresh every 5 hours up to a weekly limit, and AI credits can extend them.

Each vendor also documents accuracy limits:

OpenAI says ChatGPT can be incorrect or misleading and may sound confident when wrong. It recommends checking important information against reliable sources; access to search and research tools depends on the plan. OpenAI’s usage policies restrict tailored legal or medical advice without appropriate licensed involvement.
Anthropic says Claude can sometimes be incorrect or misleading, should not be the only source of truth, and requires careful review of high-stakes advice and cited sources. Its Acceptable Use Policy requires AI disclosure for consumer-facing chatbots and qualified review for covered advice and recommendations.
Google’s Gemini policy says Gemini should not produce factually inaccurate outputs that could cause significant real-world harm and that context matters. Google also says sources may appear within or below a response; a displayed source still needs review.

What the documentation cannot tell you

The pages do not show how the assistants handle your particular mix of European or Asian languages, Arabic or Hebrew text, politeness, taboo, humor, code-switching, or accents. They also provide no like-for-like completion times for those tasks or measurements of how usage limits behave under your workload.

Those differences can depend on the prompt, selected tools, account settings, network, and input material. The previous version of this page contained unsourced statistics and test results; those claims have been removed.

How to check it yourself

Use the same device, connection, app version, and account tier where possible. Keep each full output and do not revise the prompt between assistants.

European language performance. Give each assistant: “Translate this notice into Finnish, Polish, and Swedish, then back-translate each result into English: ‘The community hall closes before evening registration ends. The last entry is at the west door, and unpaid community members must ask staff for assistance.’” Look for preserved meaning, natural grammar, consistent terminology, and differences revealed by the back-translation. Record substantive errors and awkward phrasing.
Asian language nuance. Give each assistant: “Rewrite this sentence for a Japanese professor, a senior Korean colleague, and a senior Chinese colleague. Preserve the scheduling problem, use respectful wording suited to each relationship, and explain your choices: ‘Please move our meeting because I cannot attend at the agreed time. I am sorry for the inconvenience.’” Look for changes in directness, formality, and implied meaning. Record unsupported cultural assumptions as well as wording you would keep.
Arabic and Hebrew layout. Give each assistant: “Translate this sentence into Arabic and Hebrew, then back into English: ‘The document is in the folder beside the entrance, and the office is open after lunch.’ Keep the result readable in a right-to-left interface and flag mixed-direction text.” Look for broken word order, misplaced punctuation, incorrect script mixing, and unclear line breaks. Record each rendering problem before reading the back-translation.
Cross-cultural pragmatics. Give each assistant: “Adapt this refusal for a Japanese business email, a Brazilian Portuguese customer message, and an Egyptian Arabic workplace chat: ‘We cannot approve this request today.’ Keep the decision unchanged, make each version natural, and label any cultural assumption you cannot support.” Look for whether politeness changes the underlying decision or creates a new commitment. Record which claims are explained and which are presented as general rules.
Speed and cost efficiency. Give each: “Summarize this sentence and identify any ambiguity: ‘The library will open later on weekdays, but its Saturday hours will not change during the renovation.’” Look at completion time, completeness, tool use, and any usage warning. Record the plan and displayed monthly price, usage meter before and after, elapsed time, and any cap message. If a meter is unavailable, write “unavailable” rather than estimating a token cost.

Which rows of the comparison matter

In the Claude vs ChatGPT vs Gemini comparison matrix, read these rows in order: Free plan, Low-cost tier, Main paid plan, Heavy-use plans, Team plan, Context window in the app, How usage limits are described, Training on your chats, and Ads.

The plan rows establish access conditions and published spending limits. Context and usage rows help identify capacity constraints, but they do not predict translation quality or nuance. Training and advertising rows affect operational choices. Every populated matrix figure links to its vendor page and read date; an unconfirmed figure remains blank and marked “Not verified.”