Mongolian Speech-to-Text Leaderboard
Whisper turbo MNOperating point: Blgn94/whisper-large-v3-turbo-mn-lora (merged LoRA) on mps

Whisper turbo MN — Mongolian Speech-to-Text Benchmark

Real results from the Whisper turbo MN Batch API (enhanced operating point, language mn) evaluated across 4 Mongolian datasets and 265 audio samples. Every number below is measured — not marketing. WER, speed, and pricing are shown as-is, brutally honest.

Language:MNDiarization:noneSamples:265Run:Aug 18, 2026, 6:01 PMEndpoint:local (transformers)
06710074.3%
Accuracy

265/265 samples transcribed · 100% success rate

06710025.7%
Word Error Rate

Lower is better · across 265 samples

Headline metrics
Character Error Rate9.2%
Avg speed factor2.43×realtime multiple
Total speed factor2.43×
Avg latency / sample3.0s
Total audio processed1849.3s30.8 min

Pricing

Whisper turbo MN list pricing for batch transcription. No discounts, no negotiated rates applied — the raw per-minute rate.

Per 1k minutes
$0
batch
Per minute
$0.0000
effective
Per 1k min (these 30.8 min)
$0.00
would cost
Open source?
proprietary

Pricing source: Whisper turbo MN public pricing. Duudlaga Flow is shown for context only — this page isolates Whisper turbo MN so the number is not padded by our own product.

WER & CER by dataset

Word and character error rates per dataset. Lower is better — and these are the real Whisper turbo MN numbers, which are weak on Common Voice 24.

Common Voice 24 (MN)
Source dataset →
WER
18%
CER
6%
Shunya Labs Mongolian Speech
Source dataset →
WER
17%
CER
4.7%
Common Voice 20 (MN)
Source dataset →
WER
19.1%
CER
6.8%
Modern Voice
WER
40.1%
CER
15.7%
WERCER

Dataset summary

Aggregate accuracy, speed, and timing for each dataset.

DatasetSourceSamplesSuccessWERCERAccuracySpeedAudio (s)Proc (s)
Common Voice 24 (MN)Hugging Face →5959/5918%6%82.0%1.98×330.6s167.1s
Shunya Labs Mongolian SpeechHugging Face →6060/6017%4.7%83.0%2.9×645.1s222.3s
Common Voice 20 (MN)Hugging Face →5454/5419.1%6.8%80.9%1.99×269.8s135.9s
Modern Voice9292/9240.1%15.7%59.9%2.56×603.8s235.9s

WER vs speed — per sample

Each dot is one audio sample. The sweet spot is the bottom-left (low error, fast). Whisper turbo MN sits high on error for many Common Voice 24 samples.

Per-sample results

Ground truth shown verbatim in the Expected column. The result column highlights only the words Whisper turbo MN got wrong, in red — no strikethrough/swap gymnastics, just the mistakes.

265 rows
redwrong word in the resultplaincorrectly transcribedExpected column shown verbatim as ground truth
#AudioSampleDatasetExpected (ground truth)Whisper turbo MN resultWERCERI/D/S
1btsee_0001Common Voice 24 (MN)Гэхдээ амьсгал хураахаасаа өмнө танд мэдэж байгаагаа хэлье.Гэхдээ амьсгал хураахаасаа өмнө танд мэдэж байгаагаа хэлье.0.0%0.0%0/0/0
2btsee_0002Common Voice 24 (MN)Надад заяасан аз жаргал гэдэг ердөө гуравхан сарын хугацаатай байсан гэж үү?Надад заяасан аз жаргал гэдэг ердөө гуравхан сарын хугацаатай байсан гэж үү?0.0%0.0%0/0/0
3btsee_0003Common Voice 24 (MN)Одоо бид өөрсдөө өвчин эмгэгээсээ салахыг хичээцгээе.Одоо бид өөрсдөө өвчин эмгэгээсээ салахыг хичээцгээе.0.0%0.0%0/0/0
4btsee_0004Common Voice 24 (MN)Би бол голдуу хээрээр гэр, хэцээр дэр хийж явдаг хүн.Би бол голдуу хээрээр гэр хэцээр дэр хийж явдаг хүн.0.0%0.0%0/0/0
5btsee_0005Common Voice 24 (MN)Хан хурмаст уурлаж, Болдоггүй Бор өвгөнийг хор луугаараа ниргүүлэхээр явуулжээ.Хан хурмаст уурлаас Болдоггүй Бор өгнөө хоёр луугаар нэргүүлэхээр явуулжээ.50.0%14.3%0/0/5
6btsee_0006Common Voice 24 (MN)Алив наашаа ороод ир гээд гэртээ оров.Алив наашаа ороод ир гэээд гэртээ оров.14.3%2.7%0/0/1
7btsee_0007Common Voice 24 (MN)Өө өндөр дээдэс таны тухайд би баталж чадахгүй.Өө өндөр дээдэс таны тухайд би бодол ч чадахгүй.25.0%10.9%1/0/1
8btsee_0008Common Voice 24 (MN)Харин гурав дахь удаагаас эхлэн хүмүүсийг сонирхож эхлэв.Харин гурав дахь удаагаас эхлэн хүмүүсийг сонирхож эхлэв.0.0%0.0%0/0/0
9btsee_0009Common Voice 24 (MN)Та нар очингуутаа шөл л өгч үз.Та нар очингуутаа шүл л өгч биз.28.6%10.0%0/0/2
10btsee_0010Common Voice 24 (MN)Ерөөсөө литр үйлдвэрээсээ салаагүй явсан юм чинь.Ерөөсөө л үйлдвэрээсээ салаагүй явсан юм чинь.14.3%6.2%0/0/1
Page 1 of 27

Methodology

How these numbers were produced.

Provider: Whisper turbo MN (Whisper large-v3-turbo fine-tuned on Mongolian Common Voice + FLEURS (LoRA merged into the base weights) — the most recent Mongolian adaptation of a current-generation Whisper.). Models with this badge were trained on data that overlaps the benchmark corpora (Mongolian Common Voice — 113 of 173 samples — or the public Shunya Labs corpus). Scores on overlapping corpora are inflated by memorization; judge these models by the per-dataset table on their details page, especially the corpora they were NOT trained on.

Endpoint local (transformers). Language mn. Operating point Blgn94/whisper-large-v3-turbo-mn-lora (merged LoRA) on mps. Diarization none.

Datasets: Common Voice 24 (MN), Shunya Labs Mongolian Speech, Common Voice 20 (MN), Modern Voice — 265 samples, 1849.3s of audio total.

Dataset source URLs:

Metrics: WER and CER are computed with a standard word/character Levenshtein alignment, normalized for case and punctuation. Accuracy = 100 − WER. Speed factor = audio duration ÷ processing time (× realtime). All requests are real Whisper turbo MN Batch API calls, not cached or simulated.

Diff highlighting: The result column aligns to the ground truth and colors every substitution and insertion red. Deletions (words missing from the result) are not shown in the result column — the Expected column already holds the full ground truth as-is.

Generated by the Whisper turbo MN Benchmark Runner · Whisper turbo MN Batch API v2 · run Aug 18, 2026, 6:01 PM