Estimated reading time: 12 minutes
Search for apps for English language learners and almost everything you find is written for an adult teaching themselves English on a commute. Those lists are not wrong, they are answering a different question. A district serving English Learners is not choosing a personal study habit; it is choosing something that works inside a school day, aligns to state ELD standards, survives a privacy review, and produces evidence a teacher can act on. This guide sorts the category by what each type of app actually does, says plainly which ones move proficiency and which ones do not, and covers the procurement and funding questions the consumer lists never touch.
Table of contents
Executive Summary
Apps for English language learners fall into five categories: vocabulary and grammar practice, reading and leveled text, translation and access, teacher productivity, and speaking practice. The first four are well served by many products and are largely interchangeable. The fifth is thinly served, because holding a responsive spoken exchange with one student does not get cheaper at scale the way generating or scoring text does. For a K-12 EL programme the decisive question is not which app is best overall but which category addresses the constraint the programme actually has, which is almost always oral production.
Key Takeaways
- Consumer app rankings optimise for adult self-study, not K-12 programme outcomes.
- Four of the five app categories are crowded and broadly interchangeable; one is not.
- A student can spend twenty minutes in a vocabulary app and produce zero spoken words.
- Standards alignment, FERPA and COPPA review, and rostering matter more than feature lists.
- Title III can fund educational technology as a supplemental service.
- Free authoritative material should be exhausted first: see our English language learners resources guide.
The Five Categories of Apps for English Language Learners
Sorting the market by job rather than by brand makes the decision tractable, because it shows immediately where a district already has coverage and where it does not.
| Category | What it does | Market depth | What it cannot do |
|---|---|---|---|
| Vocabulary and grammar | Drills words, forms and structures with spaced repetition | Very crowded | Produce extended language |
| Reading and leveled text | Serves texts matched to reading level, often with glossing | Crowded | Build oral fluency |
| Translation and access | Removes a comprehension barrier in the moment | Crowded, largely commoditised | Teach English; it substitutes for it |
| Teacher productivity | Generates materials, differentiates texts, drafts communications | Very crowded since 2023 | Add student practice time |
| Speaking practice | Elicits spoken language and responds to it | Thin | Replace a teacher’s judgement |
Quick answer: most apps for English language learners do one of four things that many products do well and cheaply. Only the fifth category, responsive speaking practice, addresses the constraint that limits most EL programmes, and it is the category with the fewest genuine entrants.
Why the Consumer App Lists Mislead District Buyers
The apps that dominate published rankings are built for a motivated adult choosing to study a language. That user sets their own pace, needs no standards alignment, generates no student data, and answers to nobody. Almost none of that transfers to a school.
Three differences matter enough to disqualify most consumer apps from a district roster.
- Data and privacy. A consumer app collects data under a consumer privacy policy. A school needs a product that operates as a school official under FERPA, complies with COPPA for students under 13, and will sign a state data privacy agreement. Most consumer apps will do none of these. See FERPA and AI for what to verify.
- Standards alignment. A district is accountable for progress against WIDA, ELPAC, TELPAS or its state equivalent. An app that teaches restaurant vocabulary well does not move a proficiency level that measures academic language.
- Evidence for the teacher. A consumer app reports a streak. A programme needs to know which students are progressing on which domain, and to see it before the annual assessment rather than after.
How to Evaluate an App for a K-12 EL Programme
- Name the constraint first. Look at last year’s domain scores. If speaking is the lagging domain, a vocabulary app will not move the overall level regardless of how good it is.
- Check standards alignment against your state’s framework, not against a generic CEFR mapping.
- Run the privacy review in parallel, not after selection. FERPA, COPPA, and your state’s student data privacy agreement.
- Ask what evidence it produces and whether a teacher can read it in under a minute.
- Confirm rostering and accessibility: does it integrate with your SIS, and does it meet WCAG expectations?
- Settle procurement before the pilot. A cooperative contract usually removes the RFP entirely; see cooperative purchasing.
District Benchmark: Counting Spoken Words, Not Minutes on Task
Usage dashboards report minutes, and minutes are the wrong unit for a language programme. The unit that predicts oral proficiency is words produced.
Take a student who spends twenty minutes in a well-designed vocabulary app. They may see two hundred items, tap through them accurately, and finish with a clean progress bar. The number of words they said aloud is zero. Now take the same twenty minutes in an app that requires spoken responses: at a conversational rate, that student produces somewhere in the region of six hundred to a thousand spoken words.
Across a school year of three sessions a week, the difference is not marginal. It is the difference between roughly zero and something on the order of a hundred thousand words of oral production per student. Both students show the same twenty minutes on the district’s usage report. Only one of them practised the domain the state test will measure individually.
The practical instruction is simple: when a vendor shows engagement data, ask what the student produced, not how long they stayed.
Why the Speaking Category Is Thin
The Speaking Time Gap is the shortfall between the responsive spoken practice an English Learner needs and what one teacher can supply across a full class. It explains the shape of this market better than any feature comparison.
The arithmetic, from the supply side. Generating text, scoring a multiple-choice item or leveling a passage costs essentially nothing per additional student, which is why hundreds of products do those things and why they compete on price. Holding a responsive spoken exchange with one student at a time does not behave that way: it has to listen, interpret and answer in the moment, for each student separately. That is why the category has few genuine entrants, why schools cannot staff their way around it, and why it is the one place where a purchase changes what a programme can physically do rather than making existing work faster.
It is also why the state assessments themselves administer speaking one student at a time, from kindergarten through grade twelve. The test design already concedes the constraint; the instructional day is simply not funded to match it.
Funding Apps for English Language Learners
| Funding source | Eligible uses | How to access |
|---|---|---|
| Title III, Part A | Supplemental EL technology, materials, professional development | Formula grant through the state education agency |
| Title I, Part A | Academic support in high-poverty schools | Formula grant through the state |
| State EL funding | Varies: LCFF in California, the Bilingual Education Allotment in Texas | State formula |
| Cooperative contract | Purchasing without a separate RFP | Cooperative purchasing |
Title III is supplemental by law, which makes educational technology a natural fit: it funds capacity a district is not otherwise obliged to provide. What it cannot fund is the core required programme.
Common Mistakes District Leaders Make
- Buying from a consumer ranking written for adult self-study.
- Adding a fifth tool in a category already covered while the lagging domain goes unaddressed.
- Measuring minutes rather than production.
- Treating translation as language instruction. It removes a barrier; it does not build English.
- Running the privacy review after selection, which is how pilots get cancelled in week three.
- Assuming a cooperative vetted the product. It vetted the contract, not FERPA.
Recommended Actions
Immediate (this month): Inventory the apps already in use by category. Most districts discover three products doing the same job and none doing the fifth.
Medium-term (this year): Pull last year’s domain-level assessment data, identify the lagging domain, and only then shortlist.
Long-term (strategy): Report student production, not usage minutes, in the annual EL review, alongside reclassification rate.
Questions District Leaders Should Ask
- Which of the five categories do our current apps cover, and which are duplicated?
- What was our lagging domain last year, and which app addresses it?
- How many words does a student produce in a typical session, spoken or written?
- Has every app on our roster signed a data privacy agreement?
- Can a teacher read a student’s progress in under a minute?
- Are we funding these from Title III, and is the use genuinely supplemental?
Frequently Asked Questions
What are the best apps for English language learners?
There is no single best app, because the categories do different jobs. For a K-12 programme the right question is which category addresses your lagging domain. Vocabulary, reading, translation and teacher-productivity apps are numerous and broadly interchangeable; responsive speaking practice is the thin category and usually the constraint.
Is Duolingo suitable for English Learners in schools?
It is built for motivated self-directed learners and works well for that. In a district context the gaps are standards alignment to state ELD frameworks, the evidence a teacher needs at the domain level, and the data agreements a school requires. Some districts use it as optional extra practice rather than as programme instruction.
Are there free apps for English language learners?
Yes, and free authoritative material should be exhausted before any purchase. Colorín Colorado, WIDA and NCELA publish substantial free resources; our English language learners resources guide lists them by what each one does.
Do these apps work for elementary students?
It depends on the category. Reading and vocabulary apps often assume independent literacy, which younger newcomers do not yet have. Speaking-based practice reaches students earlier, because oral language develops before literacy does. See oral language development.
What should we verify before putting an app on student devices?
That the vendor operates as a school official under FERPA, complies with COPPA for students under 13, will sign your state’s data privacy agreement, and meets accessibility expectations. Confirm all four before a pilot, not after.
Can Title III pay for an app?
Yes, where the use is supplemental to the required EL programme. Educational technology is an allowable use of Title III funds; what it cannot do is replace services the district is already obliged to provide.
Related Resources
- English Language Learners Resources: The Free List
- Technology for English Learners
- AI for English Learners
- FERPA and AI
- What Can Title III Funds Be Used For?
Conclusion
The market for apps for English language learners is deep in four places and thin in one, and the thin one is where most programmes are actually constrained. A district that sorts the category by job, checks its own domain-level data before shortlisting, and measures production rather than minutes will make a better decision than any ranking can make for it. The consumer lists are answering a question a school did not ask.
Sources: NCELA, U.S. Department of Education; National Center for Education Statistics, English Learners in Public Schools; WIDA Consortium.
The Category a District Cannot Staff Around
Every other category on the list has a human fallback. If a district has no vocabulary app, a teacher can make word cards. If it has no leveling tool, a teacher can find easier texts. Slower, but possible.
Speaking has no fallback, because the fallback is an adult sitting with one student while the other twenty-four wait. That is why the shelf of tools can look full and the oral scores still sit flat: the missing item was never a material, it was time.
Telo AI works in that category, in English plus Spanish and French for dual language settings, and the useful thing for a district evaluating the market is less the brand than the test: ask any app on your shortlist how many words the student said out loud.
See what speaking practice looks like at classroom scale: https://mytelo.ai/how-telo-works/
Sorted that way, the buying decision stops being a comparison of feature lists and becomes a question about the one input a school day cannot manufacture on its own.
Found this guide useful?
Preferred sources show up more often in your Google results, including AI Overviews.