Front Live Translation Explained (2026): Inbound & Outbound
If your customers write in languages your agents don't speak, every conversation starts with a delay — someone pasting text into a separate translator, guessing at nuance, and pasting a reply back. Front's live translation feature closes that gap inside the inbox. It detects the language of a message you receive, offers to translate it into your own language with one click, and lets you draft a reply in the language you're comfortable with and translate it out before you send. This guide covers how inbound and outbound translation actually work, where the controls surface in the conversation and the composer, which channels and languages are supported, the daily limit and settings you should know about, and the honest point where a translate button stops being enough and a reasoning layer takes over.
What "live translation" means in Front
Front's translation feature is powered by AI, and it runs in two directions. Per Front's help article on translating messages using AI, the feature automatically detects the language of a message and can translate it into your preferred language — and it works on both inbound and outbound messages.
The important design detail: a translation in Front is shared, not private. When you translate an inbound message, the translated text appears in a banner above the original and stays visible to every teammate who opens that conversation. That fits Front's whole model — the shared inbox is collaborative, so a translation one agent triggers becomes context the whole team can see rather than something living in one person's browser tab.
Translation is one of Front's AI features, and it's enabled the same way the others are: from the Front AI tab in company settings. Once it's on, the controls appear in-line where you're already working — you don't leave the conversation.
Inbound translation: detect and read
Inbound is the half you'll use most, because it's the half where a language barrier actually blocks you: a customer writes in French, German, or Japanese and you need to understand what they're asking before you can help.
How it surfaces. When Front detects that an incoming message is in a language other than your preferred one, it shows a "[Language] detected" banner on the conversation. Clicking Translate produces the translation in a banner above the original message, and an arrow icon lets you collapse or expand it so you can flip between the original and the translation without losing either.
Because the translated text is attached to the conversation, a teammate who picks the ticket up later sees the same translation you did — no re-doing the work. That's the quiet advantage of translation living in a shared inbox rather than in a personal plugin.
Outbound translation: draft, then translate before sending
Outbound is the other half, and it's where the composer comes in. The workflow Front recommends is deliberately human-in-the-loop: write your reply in the language you're fluent in, then translate it to the customer's language before you send, so you're never firing off text you can't read back.
In the composer, the AI-assist row exposes a Translate control alongside the other assist actions (Adjust tone, Fix grammar, Expand, Shorten). Open it and you pick a target language from the dropdown; Front rewrites your draft in that language in place, and you send it like any other reply.
The dropdown above is the composer Translate control on a fresh compose draft — the same control you get when replying (the shared-inbox conversation list was throwing a transient backend error at capture time, so this was shot from a new compose window; the composer and translation UI are identical either way). The visible list — Chinese Simplified and Traditional, Danish, Dutch, English, Finnish, French, German, Italian, Japanese, Korean, Polish, Portuguese, Romanian, Russian, Spanish, Swedish, Turkish, and more — is a good sense of the coverage a support team gets out of the box.
Supported channels, languages, and settings
Channels. Translation isn't email-only. Front documents it for email channels and chat-like channels — SMS, WhatsApp, Facebook Messenger, Front Chat, and custom channels. So a global team running support across email and messaging gets the same detect-and-translate behaviour everywhere.
Languages. Front's composer is built to handle far more than the translate dropdown alone. Per Front's reference on languages supported in messages and composer, Front is committed to displaying any inbound language, and the composer supports both left-to-right and right-to-left scripts — Hebrew, Arabic, Persian, Urdu — with automatic direction detection (start typing in an RTL language and the paragraph flips to right-aligned) plus a manual direction toggle in the formatting toolbar. One honest caveat from that same doc: tables stay left-to-right and don't mirror under RTL.
Settings. Behind a gear icon, translation has two controls worth knowing:
- Always translate to — sets the output language translations default to.
- Never translate — excludes chosen languages from detection. Set English here, for example, and the translate banner won't appear on English messages, so agents aren't prompted to translate what they already read fine.
The limit. There's a real rate limit to plan around: each teammate account gets 1,500 translation requests per day, and Front notes this is separate from its other AI features' limits. For most agents that's effectively invisible; for a very high-volume multilingual queue, it's a number worth knowing exists.
Where it fits — and where it stops
Let's be fair about what native translation is and isn't. Front's built-in translate is genuinely useful: it's included across Starter, Professional, and Enterprise plans (and the legacy Growth and Scale tiers), it's one click, and because translations are shared, the whole team benefits from one agent's action. For reading an inbound message and shipping a translated reply, it does exactly the job.
But notice the ceiling. Translation converts language; it doesn't resolve the ticket. A translated French message is now a readable English one — the agent still has to understand the request, look up the customer's order, and compose the answer. And the outbound flow is human-in-the-loop by design: an agent writes every reply, then translates it. That's the right default for accuracy, but it means translation speeds up a conversation without reducing how many conversations a person has to touch.
There's also a coverage seam. Front's AI translate covers a solid set of languages, but teams needing very broad language coverage or fully automatic, real-time two-way translation across dozens of languages often reach for a marketplace add-on like Lokalise Messages for Front, which uses neural machine translation across 100+ languages — an extra tool and an extra cost, layered on top.
This is exactly the seam an AI agent layer is built for. The broader category of AI agents for customer service exists to do the reasoning a translate button can't. Macha is one such layer, and it runs on top of the Front you already use through the live Macha–Front connector — it does not replace Front, your inboxes, or Front's own translation. The division of labour is clean: Front's translate makes a foreign-language message readable; a Macha agent then reads the intent (in any language, without a separate translate step), pulls the customer's real order or account status through a custom tool that turns your REST API into something the agent can call, and drafts or sends a grounded reply in the customer's language. If you want the mechanics, connecting Front to Macha to route conversations to AI walks it through, and Macha's credits are consumed per AI action — never per resolution — because translating, reasoning, and answering carry different costs and it's honest to price them separately. For how the plan tiers themselves gate features, see Front pricing explained.
FAQ
Is Front's live translation free / which plans include it? Front's AI Translate is available on the Starter, Professional, and Enterprise plans, plus the legacy Growth and Scale tiers. It's enabled from the Front AI tab in company settings. Very broad or fully-automatic multilingual coverage (100+ languages) generally requires a marketplace add-on such as Lokalise Messages for Front.
How do I translate an inbound message in Front? When Front detects a message in a language other than your preferred one, it shows a "[Language] detected" banner on the conversation. Click Translate and the translation appears in a banner above the original, with an arrow to collapse or expand it. The translation stays visible to every teammate who opens the conversation.
How do I translate my reply before sending? Draft your reply in your own language, then open the Translate control in the composer's AI-assist row, pick the target language from the dropdown, and Front rewrites your draft in that language in place before you send.
Which channels and languages does Front translation support? It works on email plus chat-like channels — SMS, WhatsApp, Facebook Messenger, Front Chat, and custom channels. The composer handles left-to-right and right-to-left scripts (including Hebrew, Arabic, Persian, and Urdu) with automatic direction detection, though tables stay left-to-right. Each teammate can make up to 1,500 translation requests per day, separate from other AI features.
Can I add AI that answers in the customer's language without replacing Front? Yes. An AI agent layer like Macha connects to Front as a live connector and runs on top of your existing inboxes — it doesn't replace Front or its translation. The agent reads intent in any language, pulls real customer data through a custom tool, and drafts or sends a grounded reply in the customer's language.
Ready to turn "translated and understood" into "actually answered"? Start a free trial of Macha and connect it to your Front in minutes.
Add AI agents to your Front
Macha resolves tickets end to end on Front — no migration, no code.
Zendesk
Freshdesk
Gorgias
Front
Shopify
Stripe
Slack
Notion
Google Workspace
Confluence

