Blending is LingoBlend's core idea: paste any text in your language, and a chosen percentage of its words come back in the language you're learning, woven into sentences you already understand — the diglot weave method with an AI doing the weaving. For two years, one AI call did everything: it read your text, decided which words to translate, translated them, and wrote the whole passage back out. That worked — but it bundled three very different jobs into one, and each job deserved a better specialist.
The problem with letting one AI pass do everything#
When a single model output carries the entire blended text, three things quietly compete:
Selection is a policy question, not a language question. Which words should be blended depends on things the AI can't know from the text alone — which words you've already mastered, which are due for review, how evenly the blend should spread across the page. A language model guesses; a selector we control can simply be told.
Re-emitting the whole text invites whole-text mistakes. If the model writes out every word of your document, every word is an opportunity for error — a dropped sentence, a duplicated phrase, a translation glued to the wrong spot. Our old pipeline carried a whole chain of repair stages that existed mostly to fix damage this architecture made possible.
Judgment gets diluted. The model's real linguistic skill — knowing that "gave up" is one unit of meaning, that "I would love" must become me encantaría with the pronoun restructured — was spent alongside bookkeeping work it should never have owned.
The new division of labor#
Blend Engine v2 splits the job along its natural seam:
| Job | Old engine | New engine |
|---|---|---|
| Choosing which words to blend | AI, in the same pass | LingoBlend's own selector, informed by the vocabulary graph and your personal mastery |
| Translating in context | AI | AI — with the full text as context, unchanged |
| Handling idioms & fixed expressions | Implicit, inconsistent | Explicit: the whole expression, or nothing |
| Assembling the final text | AI re-writes everything | Your original text stays intact; only the chosen words change |
| Catching errors | Repairs after the fact | Verified against the source before display |
Because selection is now ours, it follows principles we can actually state and test: pick vocabulary that's genuinely worth learning, spread it through the whole text rather than clumping it in the first paragraph, and — with personalization — skip what you've already mastered. The AI's job narrows to the one thing it's uniquely good at: translating those words correctly in the context of your full text.
Why a smaller job made the AI smarter#
Here's the counterintuitive part. Giving the AI a narrower job made its output better, because we could finally be explicit about the linguistic judgment we want from it:
- Idioms and fixed expressions come back whole. If a chosen word sits inside an idiom, phrasal verb, or fixed expression, the whole expression is translated — "kicked" inside "kicked the bucket" comes back as the full idiom, and tapping it in the reader saves the idiom's dictionary form, which is what's actually worth learning.
- Natural translations keep their neighbors. When the right translation swallows the words around it, they go with it. "Write to me" becomes Escríbeme as one unit — never Escríbeme with a stranded "to me" behind it.
- Words that don't translate cleanly are left alone. A name, wordplay, a word whose isolated translation would mislead — better untranslated than wrong. An untranslated word costs nothing; a wrong translation teaches you something false.
In the benchmark that mattered most to us — 93 trap sentences containing idioms, phrasal verbs, and separable verbs across all 17 source languages — the new engine expanded 65 of them into correct whole-expression translations, against 21 for the old engine. Cross-language pairs produced some of the best results: French posé un lapin rendered into German as sitzen gelassen (stood up), where the old engine offered Kaninchen — a literal rabbit.
Verified before you see it#
Because the final text is no longer written out by a model, we can guarantee properties the old engine could only hope for: every character of your original text survives, each translation lands exactly where its source word was, and a whole family of subtle errors — echoed or repeated "translations", grammar accidentally marked twice across the two languages — is caught before display. If a word fails verification, it simply stays untranslated — the failure mode is invisible instead of embarrassing.
One more benefit fell out for free: speed. Since the AI no longer generates your entire document — only the translated words — blends complete in roughly half the time, with the slowest cases improving the most. The blend you used to wait ~6 seconds for now arrives in ~3.
Everything else you know is unchanged: tap any blended word for its grammar context and save the base form or the exact form you met, re-blend any text at a new percentage, and read in the same paginated reader.