Offline Translation App Index: What 9 Apps Publish Before You Travel

Seven of the nine translation apps in this study publish a general offline claim. Only four clearly name at least one richer travel mode, camera, voice or conversation, as working offline. Google Translate and Microsoft Translator explicitly document offline camera translation. Apple Translate and iTranslate explicitly document offline conversation or voice translation for supported downloaded languages.

The difference matters because "offline" is not one feature. An app may translate typed text without a signal while its camera, microphone or two-way conversation tool still needs the internet. Another app may offer an offline dictionary without offering full sentence translation. A traveler who downloads one language pack and assumes every button will work can discover the gap at a menu, ticket machine or hotel desk.

Study result: choose the app by the task that must work without data. Download the required languages, switch on airplane mode and test typed text, camera and conversation separately. Then keep a live data option for unsupported modes, current context and a second translation route.

Download the offline translation app CSV or open the JSON dataset. The JSON contains the fixed sample rule, classification method, twelve target keywords, ten competing pages, ten public traveler questions, first-party source notes and a SHA-256 checksum for the source file. Every CSV row has its own normalized checksum.

Which translation setup fits your trip?

Why the word offline is not enough

Translation apps combine several systems that look like one feature on the screen. Typed text translation needs a language model. Camera translation also needs text recognition. Voice translation adds speech recognition and often audio playback. Conversation mode must listen, decide when a speaker has finished, identify a language and translate both directions quickly enough for a natural exchange.

Some of those systems can run on the phone after a model download. Others call a remote service. The split can also change by language, operating system and app version. A provider can truthfully advertise offline translation even when only typed text works without internet.

This study therefore uses four separate questions:

1. Does the app publish a general offline translation claim?
2. Does it explicitly document typed text translation offline?
3. Does it explicitly document camera translation offline?
4. Does it explicitly document voice or conversation translation offline?

A nearby list of online features does not answer the offline question. If a page says the app has camera translation and also says the app has offline translation, we do not combine those statements unless the source explicitly connects them. This conservative rule prevents a marketing layout from becoming a capability claim it never made.

Strongest practical finding: the more demanding the task becomes, the smaller the documented field gets. Seven apps publish a general offline claim, five explicitly name offline typed text, two name offline camera use and two name offline voice or conversation use.

Four apps publish a richer offline mode clearly

Google Translate, Apple Translate, Microsoft Translator and iTranslate receive the highlighted rows in the explorer because they explicitly connect camera, voice or conversation to offline use. The highlight is about documentation, not translation accuracy. We did not ask native speakers to score output, and we did not run controlled tests across language pairs.

Google Translate clearly documents offline camera use

Google's offline language guide says downloaded languages can be used without an internet connection. Its camera help separately explains that camera translations can work offline after the needed languages are downloaded.

That makes Google one of two apps in the sample with an explicit first-party offline camera statement. The documentation still limits the claim to supported downloadable languages. It does not say that every language or camera mode works offline on every device.

For a trip, download both sides of the language pair when the app requests them. Photograph or point the camera at a sample menu, street sign and printed booking page while the phone is in airplane mode. A successful English to Spanish text test does not prove that Japanese camera translation or a spoken exchange is ready.

Apple documents offline conversations with downloaded languages

Apple's Translate settings guide says both the input and output languages must be downloaded for use without internet. It also offers an On-Device Mode and warns that offline translations may be less accurate than online translations.

Apple's conversation guide goes further. It says conversations work with downloaded languages when the phone has no internet connection or when On-Device Mode is enabled. This is the clearest explicit offline conversation statement in the sample.

Apple also publishes a camera translation guide, but the checked camera page does not separately say that camera translation works offline. We therefore count offline conversation, not offline camera, for Apple. That may feel strict, but the purpose of the index is to show what a traveler can verify from the documentation without joining separate claims.

Microsoft publishes the clearest feature boundary

Microsoft Translator's Android help explicitly lists text translation, camera translation, phrasebooks and pinned translations as available offline after the necessary packs are downloaded. It also says speech conversations, the single microphone and split-screen microphone are not available offline.

That detail earns Microsoft the documentation clarity award in this study. It tells the traveler what to download, what works and what does not. A person trying to scan a printed sign has a documented offline path. A person planning a spoken exchange knows to arrange mobile data or another app.

🏆 Clearest offline documentation: Microsoft Translator. Its checked help page separates offline text and camera support from online-only voice and conversation modes in one place.

iTranslate publishes offline text and limited voice counts

iTranslate's offline feature page publishes offline text translation in 41 languages and offline voice translation in five languages. It also lists offline phrasebook access. The feature comparison shows offline mode as part of iTranslate Pro.

The numbers make the travel decision clearer. A traveler cannot use the broad online language count as the offline voice count. The exact five-language set and current subscription entitlement should be checked in the app before purchase and before departure.

iTranslate also offers camera translation, but the checked offline page lists camera as a separate Pro feature rather than explicitly placing it inside offline mode. The dataset therefore does not count offline camera support.

Three apps claim offline translation without mapping every mode

Papago, Lingvanex and the broader parts of several product pages show why a general offline claim needs another question. They may be useful apps. The evidence simply does not support filling in every mode on their behalf.

Papago publishes offline translation but not the mode split

The developer-supplied Papago listing names text, image, voice, offline and conversation translation as separate features. It also refers to downloadable offline translation content.

The listing does not explicitly say whether offline mode covers typed text only or also covers image, voice and conversation. This study records the general claim as Yes and the three specific modes as Scope not specified.

That distinction is especially relevant for travelers choosing Papago for Korean, Japanese or Chinese. Install the current app, download the available offline content and run three airplane-mode tests. Type a sentence. Scan a printed sign. Try a two-way spoken exchange. Keep the result for the exact language pair, phone and app version you will carry.

Lingvanex publishes a broad mobile offline claim

Lingvanex's mobile page says the mobile translator works offline and instructs users to download a language. The same page lists text, voice, files and images as app features.

It does not clearly map each listed mode to offline use. The study counts the general offline claim but does not mark camera or voice as explicitly documented offline. This is not a negative judgment about the app. It is a reminder that broad product copy cannot answer a mode-specific travel question without a connecting statement.

The safe preparation is the same as for Papago: install, download and test each required input in airplane mode. Confirm any payment or account requirement before the refund window closes.

Yandex defines offline mode as typed text

Yandex Translate's offline guide is more specific. It says manually entered text can be translated offline and that voice input and dialog mode require internet. English and Russian offline dictionaries are included, while other languages must be downloaded.

Yandex is counted among the seven apps with a general offline claim and among the five with explicit offline text. It is not counted for camera, voice or conversation. Its strength is clarity: a traveler knows it is a typed-text fallback and can plan a connected tool for the other tasks.

Two entries do not support a full offline assumption

Reverso and DeepL illustrate two different documentation outcomes. Reverso publishes limited offline material. DeepL's checked mobile pages do not publish offline use.

Reverso publishes offline dictionaries, not every translation mode

Reverso's mobile page advertises offline dictionaries. Its product material also names camera and voice translation, while another first-party page lists history, favorites and flashcards among offline Premium features.

The offline label therefore applies to dictionaries and saved learning material in the checked sources. It does not explicitly extend to live camera or voice translation. A traveler can still benefit by downloading a dictionary and saving common phrases, but should not treat that preparation as proof that a photographed menu will translate without data.

This is a useful category in its own right. A prepared phrasebook can be faster than any live translator for an address, allergy, medication, dietary restriction or request for help. The study labels the capability Limited rather than No.

DeepL's checked mobile pages do not publish offline use

DeepL's mobile page publishes text translation, text to speech, speech to text, camera capture and uploaded photo or document translation. The developer's Android listing gives a similar feature set. Neither checked page advertises an offline mode.

The classification is Not published on the checked mobile pages. It is not a claim that no cached screen, saved favorite or unlisted behavior can ever appear. It means a traveler should not build a no-signal plan around an offline capability that the current public pages do not promise.

DeepL can still be part of a connected travel setup. Save critical translations before leaving reliable service and keep a documented offline app for dead zones. Use live mobile data when you want DeepL's current online tools away from hotel Wi-Fi.

Where Roaming eSIM has the connectivity edge

Offline language packs are the right answer when the phone has no usable network. They do not remove the value of live data. Connected translation can access modes that a provider excludes from offline use, handle a language pair that has no downloadable pack, translate a web page and combine text with current place information.

A route-matched Roaming eSIM plan can supply mobile data for those connected tasks on a compatible, carrier-unlocked phone. It can also keep maps, ride booking, transit changes, hotel messages and support available on the same trip. The selected package card remains the source of truth for coverage, network, allowance, validity, hotspot terms, activation and top-up eligibility.

🏆 Roaming eSIM connectivity edge: offline packs cover the dead zone. Roaming eSIM covers the connected modes before and after it, without tying the traveler to airport or hotel Wi-Fi. The best setup uses both rather than pretending either one solves every translation problem.

Roaming eSIM is not a translation engine and cannot make an unsupported language available inside a third-party app. It also cannot guarantee signal in a tunnel, remote road or shielded building. That boundary is why the study recommends a tested offline pack alongside mobile data.

For the rest of the phone setup, download an area in the Offline Maps Travel Index, check how much mobile data travel apps may use, confirm when to install a Roaming eSIM and review whether a plan includes calls or SMS.

How to prepare a translator before departure

1. List the tasks, not only the languages

Write down what the phone must do without a connection. Reading a printed menu needs camera translation. Showing a hotel clerk a typed sentence needs text translation. Understanding a spoken reply needs speech recognition or conversation mode. Looking up a saved allergy phrase may need only a phrasebook.

This prevents a basic offline text claim from being mistaken for a complete travel setup. It also helps avoid paying for a feature that the trip does not need.

2. Download every required language component

Some apps need one language pack. Others need both source and target languages. Microsoft and Apple explicitly tell users to download both sides for offline use. Storage requirements can grow across a multi-country trip, so download while reliable Wi-Fi and charging are available.

Open the language settings after download and confirm that the app marks each language as available offline. Do not rely on a progress notification that may have been interrupted when the screen locked.

3. Run an airplane-mode test for each input

Turn on airplane mode, then keep Wi-Fi off. Test the exact tasks:

• type a common travel sentence in both directions;
• scan a clean printed menu or sign;
• import a saved photograph if that matters to the trip;
• speak a short phrase and check whether the app captures it;
• try a two-way conversation if another person will use the phone;
• play the translated audio if pronunciation output is important;
• close and reopen the app to confirm the downloaded model remains available.

An offline test is more useful than a provider-wide label because it includes the actual device, operating system, app version and language pair.

4. Save high-consequence phrases separately

Machine translation can be wrong. Save critical information in a form that does not depend on a live translation request. Examples include the accommodation address, emergency contact, medication name, diagnosed allergy and a professionally reviewed explanation of a severe dietary restriction.

Do not use a consumer translation app as the sole authority for medical, legal, immigration or safety-critical communication. Ask a qualified interpreter or relevant professional when an error could cause harm.

5. Add live data without deleting the offline fallback

Install the Roaming eSIM and confirm that the plan covers the actual route. Keep the offline languages in place after mobile data starts working. A busy cell, underground station or remote area can still interrupt online translation.

On a dual-SIM phone, select Roaming eSIM for mobile data and review whether automatic data switching should be disabled. The exact menus vary by device. The setup and support hub provides current Roaming eSIM installation paths and contact options.

What travelers are asking about offline translation

Public Reddit discussions were used to identify recurring questions, not to prove app behavior. The official sources and classification rules determine every app row.

Does downloading a language make every feature work offline?

No. The checked documentation shows separate boundaries. Microsoft supports offline text and camera but not its documented voice or conversation modes. Yandex limits offline use to manually entered text. iTranslate publishes different language counts for offline text and offline voice.

Can a phone translate a restaurant menu without data?

Google Translate and Microsoft Translator explicitly document offline camera use for supported downloaded languages in the checked pages. Camera recognition can still struggle with handwriting, stylized fonts, glare, low light or a language that is not downloaded. Test a representative printed menu before departure.

Which app documents an offline two-way conversation?

Apple says conversations work with downloaded languages on supported iPhones. iTranslate publishes offline voice translation in five languages. This study does not claim that both apps have equal conversation design, accuracy or language coverage.

Should I trust camera translation for an allergy?

No consumer app should be the only control for a severe allergy. Save a professionally reviewed allergy card in the local language, show it to staff and confirm ingredients and cross-contact directly. Camera translation can help read a label, but it cannot inspect a kitchen process.

Do I need mobile data if I have offline packs?

Offline packs can cover specific tasks without data. Mobile data is still useful for online-only modes, unsupported languages, website translation, current place context, transport changes and a second translation service when the first result is unclear.

Can I download the pack after landing?

Yes, when the app allows it and the phone has a reliable connection, but this is a weak primary plan. Airport Wi-Fi may require a portal, hotel Wi-Fi may not be available yet and mobile service may be the reason the translator is needed. Download and test before departure.

Does offline mean private?

Offline processing can reduce the need to send a translation request to a remote server, but the word offline does not describe every privacy behavior of the app. Account services, analytics, crash reports, updates and other features may still connect later. Read the current privacy policy for a separate privacy decision.

Is a paid app automatically better offline?

No. Payment and capability are separate fields. iTranslate places offline mode in Pro and publishes useful mode counts. Free Google and Microsoft documentation also explicitly covers offline camera use. Choose by the exact language and task, then test it.

What if the camera translation gives nonsense?

Capture a still image in good light, crop to the relevant text and compare a second translator when data is available. Ask a person for help when the decision matters. Do not guess at medical instructions, legal notices, allergens or transport restrictions.

Should I carry a dedicated translator device?

This study does not compare hardware. A dedicated device may suit someone who wants a separate battery, microphone or interface. Check whether it genuinely works offline, which languages are stored locally, whether it needs a subscription and how models are updated.

Why travelers need a mode-by-mode offline translation index

We reviewed ten search-visible pages across the target keyword set. The leading results fell into three groups: product pages for one offline translator, broad editorial roundups and commercial comparisons that recommend the publisher's own app.

Several pages correctly note that offline text, camera and voice are different. Few provide a downloadable dataset with one fixed rule, current first-party source links, per-row checksums and explicit Not published classifications. Some join a general offline claim to a nearby camera or conversation bullet without showing that the provider made that connection.

This study does not try to win the broadest "best translation app" claim. It owns a narrower decision: what nine apps currently publish about the modes a traveler expects to work when mobile data disappears.

The search strategy follows that decision:

• primary intent: offline translation app for travel;
• comparison intent: offline text, camera and conversation documentation;
• setup intent: language-pack downloads and airplane-mode testing;
• product connection: live mobile data for modes that remain online;
• citation asset: reproducible CSV and JSON with dated first-party sources.

How to reproduce the offline translation dataset

Sample selection

The fixed purposive sample contains nine mobile translation apps with current first-party documentation in English. It includes Google Translate, Apple Translate, Microsoft Translator, Naver Papago, iTranslate, Yandex Translate, Reverso Context, Lingvanex and DeepL.

The sample covers default phone platforms, large cross-platform services, travel-focused products and one widely searched connected translator. It is not based on market share, download rank or translation quality. Smaller offline-first apps can have strong features but are outside this fixed comparison.

Source hierarchy

We used provider help centers and official product pages first. A developer-supplied Google Play listing was used when the provider's accessible help did not publish the needed mobile feature description. Search snippets, Reddit comments and competitor articles were not used to classify app capability.

Every external source link in the article is marked nofollow under the resource-page policy. The CSV preserves the primary source for each row. The JSON preserves additional sources and the reason for each classification.

How each offline translation mode was counted

A general offline claim is Yes only when the checked first-party source says the mobile app works offline or supports offline translation. Limited is used when the offline material is narrower, such as dictionaries and saved study content. Not published is used when the checked current mobile pages do not advertise offline use.

Typed text, camera and voice or conversation are counted separately. A specific mode is Yes only when the source explicitly connects it to offline use. No is used when the provider explicitly says the mode requires internet. Scope not specified or Not published is used when the provider does not resolve the question.

The documentation rule is intentionally strict. It measures what a traveler can verify before the trip, not every behavior hidden inside every regional build.

Limits of this translation documentation audit

The index does not score translation accuracy, latency, pronunciation quality, handwriting recognition, privacy, battery use, model size, accessibility or language breadth. It does not install the apps or run a controlled airplane-mode lab. It does not claim that a published feature performs equally across languages.

Those limits keep the finding honest. A first-party documentation audit can answer whether a capability is clearly promised. It cannot answer how well a Japanese handwritten menu, Arabic road sign or noisy Spanish conversation will translate on a particular phone.

Reproduce or update the data

1. Open scripts/research/data/offline-translation-app-index-source.json in the public repository. 2. Review the fixed sample, classification rule and dated first-party URLs. 3. Recheck every cited page for a changed offline claim or mode boundary. 4. Update only the affected source record and preserve the wording that distinguishes No from Not published. 5. Run node scripts/research/generate-offline-translation-app-index.mjs. 6. Compare the new source-file checksum, row checksums, CSV, JSON and three WebP charts. 7. Update the article finding only when the regenerated summary changes.

The generator creates the published CSV, JSON and charts from the source file. Running it twice without source changes produces identical data files.

Source notes by app

The practical connectivity decision

Downloadable language packs reduce dependence on any network, so they belong on the phone even when the traveler buys mobile data. Live data expands the available tools, supports online-only modes and provides a second route when the offline result is unclear.

For most trips, the evidence supports a two-layer plan:

1. Prepare one app whose exact offline task is explicitly documented and tested on the travel phone.
2. Add a Roaming eSIM plan that covers the destination for connected camera, voice, web and context features.

That decision follows the data even though Roaming eSIM does not "win" inside every translation row. The product solves the connection layer. The chosen translation app controls language support and offline behavior. Keeping those roles separate produces a more reliable trip plan than treating either tool as a complete substitute for the other.

Research updates and corrections

This study was checked on August 21, 2026. We will update the existing URL when a sampled app changes its public offline claim, mode coverage, payment boundary or setup instructions. We will also correct a row when a first-party page shows that a classification no longer matches the evidence.

Send a correction with the app, platform, language pair, version and first-party source. A screenshot can help locate a change, but the published dataset will not convert an in-app observation into a provider-wide claim without enough scope to classify it accurately.

The study remains valid if a future update reduces the number of apps with explicit offline features or if a connected app improves its documentation. The finding follows the sources, not a preferred winner.