The same team that builds the TonePerfect app. The API scores Mandarin the way the app teaches it — initial, final and tone for every syllable — and adds six further language packs built on the same approach. Every response names the model version that scored it.

Plans start at $10/month for 2,500 credits; one credit scores a recording up to 15 seconds and failed requests are free. There is no free tier, but creating an account, keys and reading the docs costs nothing.

Is learner audio stored?

No. Audio is deleted after scoring and never used for training. Results are kept for up to 24 hours so a retried request returns the same answer.

Can I use it for languages other than Chinese?

Yes: English, French, Spanish, German, Italian and Portuguese are served through the same endpoint with the same JSON shape.

For developers

The TonePerfect pronunciation engine, as an API

Send a short recording and the sentence the learner was asked to read. Get back a 0–100 score for the sentence, every word and every sound — including Mandarin initials, finals and tones — in one JSON response. Seven languages, one endpoint, from $10/month.

Explore the API →Read the docs

One request

Audio, reference text and a locale. Add an Idempotency-Key header to retry safely on flaky mobile networks.

curl https://api.toneperfect.app/v1/assess \
  -H "Authorization: Bearer $TONEPERFECT_API_KEY" \
  -F "audio=@take.wav" \
  -F "text=你好世界" \
  -F "language=zh-CN"

Three levels of detail

Sentence, word and sound scores, already linked — so your app can go from “86” to “your 4th tone came out as a 2nd” in one tap.

{
  "scores": { "overall": 86, "pronunciation": 90, "initial": 91, "final": 89, "tone": 80 },
  "words": [
    { "word": "世", "pinyin": "shi4", "score": 68, "start_ms": 1260, "end_ms": 1640,
      "phones": [
        { "dimension": "initial", "expected": "sh", "heard": "s", "score": 78, "decision": "uncertain" },
        { "dimension": "final",   "expected": "i",  "heard": "i", "score": 83, "decision": "no_correction" },
        { "dimension": "tone",    "expected": 4,    "heard": 2,   "score": 62, "decision": "review" }
      ] }
  ]
}

Seven languages, the same JSON

Every pack returns sentence, word and sound scores. English and French add fluency and rhythm; Mandarin adds initials, finals and tones with sandhi handled.

Decisions, not just numbers

Every word — and every Mandarin initial, final and tone — carries a review / uncertain / no-correction decision, so your app knows when to correct and when to stay quiet.

Timestamps on every word

start_ms and end_ms let you replay the exact segment a learner needs to hear again.

Audio is never stored

Recordings are assessed and deleted, never used for training. Results expire within 24 hours.

Questions

Who builds the API?: The same team that builds the TonePerfect app. The API scores Mandarin the way the app teaches it — initial, final and tone for every syllable — and adds six further language packs built on the same approach. Every response names the model version that scored it.
What does it cost?: Plans start at $10/month for 2,500 credits; one credit scores a recording up to 15 seconds and failed requests are free. There is no free tier, but creating an account, keys and reading the docs costs nothing.
Is learner audio stored?: No. Audio is deleted after scoring and never used for training. Results are kept for up to 24 hours so a retried request returns the same answer.
Can I use it for languages other than Chinese?: Yes: English, French, Spanish, German, Italian and Portuguese are served through the same endpoint with the same JSON shape.

Get your API key →