validcube

AI visibility index · Language learning apps · snapshot 1 September 2026

Which language-learning apps the assistants recommend

35 buyer questions across 5 assistants. The score ranks apps by how often the search-grounded assistants name them.

The fourth snapshot. Captured 1 September 2026 across 5 assistants, one month after the last. The chart below tracks all four months. How this works.
Number = share of an assistant's answers that named the app (%). Score = average of the 5 grounded assistants. ▲/▼ = rank change vs last month. +/− next to Score = point change vs last month.
# App ChatGPT Claude Gemini Mistral Google AI · web Score
1 —
40
60
43
54
67
53 +9
  • ChatGPT
    40
  • Claude
    60
  • Gemini
    43
  • Mistral
    54
  • Google AI · web
    67
2 —
34
34
34
29
67
40 +3
  • ChatGPT
    34
  • Claude
    34
  • Gemini
    34
  • Mistral
    29
  • Google AI · web
    67
3 —
14
14
31
29
50
28 +4
  • ChatGPT
    14
  • Claude
    14
  • Gemini
    31
  • Mistral
    29
  • Google AI · web
    50
4 ▲3
9
14
37
23
33
23 +9
  • ChatGPT
    9
  • Claude
    14
  • Gemini
    37
  • Mistral
    23
  • Google AI · web
    33
5 ▼1
23
20
29
6
33
22 +1
  • ChatGPT
    23
  • Claude
    20
  • Gemini
    29
  • Mistral
    6
  • Google AI · web
    33
6 ▲2
9
11
31
17
33
20 +7
  • ChatGPT
    9
  • Claude
    11
  • Gemini
    31
  • Mistral
    17
  • Google AI · web
    33
7 ▲2
6
17
17
20
17
15 +3
  • ChatGPT
    6
  • Claude
    17
  • Gemini
    17
  • Mistral
    20
  • Google AI · web
    17
8 ▼2
9
14
3
26
17
14 -1
  • ChatGPT
    9
  • Claude
    14
  • Gemini
    3
  • Mistral
    26
  • Google AI · web
    17
9 ▼4
3
6
20
9
33
14 -3

Name collision: "Speak" is also a common word or a bigger brand, so matches are checked in context.

  • ChatGPT
    3
  • Claude
    6
  • Gemini
    20
  • Mistral
    9
  • Google AI · web
    33
10 ▲10
3
6
14
11
33
14 +11
  • ChatGPT
    3
  • Claude
    6
  • Gemini
    14
  • Mistral
    11
  • Google AI · web
    33
11 ▲5
3
11
17
14
17
12 +7

Name collision: "Tandem" is also a common word or a bigger brand, so matches are checked in context.

  • ChatGPT
    3
  • Claude
    11
  • Gemini
    17
  • Mistral
    14
  • Google AI · web
    17
12 ▼2
14
20
3
9
17
12 +1
  • ChatGPT
    14
  • Claude
    20
  • Gemini
    3
  • Mistral
    9
  • Google AI · web
    17
13 ▲1
3
3
29
9
0
9 +3
  • ChatGPT
    3
  • Claude
    3
  • Gemini
    29
  • Mistral
    9
  • Google AI · web
    0
14 ▼2
0
3
23
3
17
9 ±0
  • ChatGPT
    0
  • Claude
    3
  • Gemini
    23
  • Mistral
    3
  • Google AI · web
    17
15 —
3
3
9
3
17
7 +1
  • ChatGPT
    3
  • Claude
    3
  • Gemini
    9
  • Mistral
    3
  • Google AI · web
    17
16 ▼5
0
6
3
3
17
6 -5

Name collision: "ELSA Speak" is also a common word or a bigger brand, so matches are checked in context.

  • ChatGPT
    0
  • Claude
    6
  • Gemini
    3
  • Mistral
    3
  • Google AI · web
    17
17 ▲1
3
9
0
3
17
6 +2
  • ChatGPT
    3
  • Claude
    9
  • Gemini
    0
  • Mistral
    3
  • Google AI · web
    17
18 ▼1
0
11
0
11
0
5 +1
  • ChatGPT
    0
  • Claude
    11
  • Gemini
    0
  • Mistral
    11
  • Google AI · web
    0
19
0
0
3
0
17
4
  • ChatGPT
    0
  • Claude
    0
  • Gemini
    3
  • Mistral
    0
  • Google AI · web
    17
20 ▼1
0
3
6
6
0
3 -1
  • ChatGPT
    0
  • Claude
    3
  • Gemini
    6
  • Mistral
    6
  • Google AI · web
    0
21 ▼8
3
6
3
6
0
3 -5

Name collision: "Drops" is also a common word or a bigger brand, so matches are checked in context.

  • ChatGPT
    3
  • Claude
    6
  • Gemini
    3
  • Mistral
    6
  • Google AI · web
    0
22 ▼1
0
6
6
3
0
3 ±0
  • ChatGPT
    0
  • Claude
    6
  • Gemini
    6
  • Mistral
    3
  • Google AI · web
    0
23 ▼1
0
3
3
0
0
1 ±0
  • ChatGPT
    0
  • Claude
    3
  • Gemini
    3
  • Mistral
    0
  • Google AI · web
    0
24
0
0
0
3
0
1
  • ChatGPT
    0
  • Claude
    0
  • Gemini
    0
  • Mistral
    3
  • Google AI · web
    0

Google AI is a single web capture of the AI Overview for 6 of the prompts, not the full set, so treat its column as directional. Every other assistant runs the full prompt set once each.

How the numbers moved

One snapshot is a photo. The index is the film. Each monthly run adds a column here, so a climb or a slide shows up as a line, not a number you have to remember.

Score over time · top 6
JunJulAugSepDuolingo53  +9Babbel40  +3Pimsleur28  +4italki23  +9Busuu22  +1Anki20  +7

A leader, and a disruption under it

Duolingo leads the category, named in nearly half of grounded answers and ahead of every challenger. But ask specifically for an AI tutor and its grip loosens: on the best-AI-app-to-learn-a-language question, Gemini and Mistral both open with Langua, a specialist, and Google's AI Overview follows them rather than the incumbent. The generic question is Duolingo's; the AI question is not.

The assistant won't name itself

The obvious AI tutor is the assistant you're already talking to. Yet asked for the best AI-powered way to learn a language, Google's AI Overview sends the question to dedicated apps like Langua and Speak. Across the six captures it names Gemini, its own model, zero times and ChatGPT once, in passing. It points you elsewhere rather than to itself, the mirror image of the self-preference I looked for in the AI index.