Speech to text

Распознавание карельской речи (Mihaj)

About two dozen Whisper, wav2vec2 and WavLM fine-tunes for Karelian, including recordings where the speaker switches from Karelian to Russian. The model cards are empty, so quality must be checked on your own recordings.

Developer
Mihaj Dolgushin, not disclosed
First release
Feb 2024
Latest release
Nov 2025
Sizes
from 300M to 1B across the different models
License
Commercial use with conditionsno license is stated on the model cards — check with the developer
Russian
Supported
Running
On your own serverAlso runs without a GPU
Industries
Media and production, Education, Science and research, Public sector
Deploy on your server Hugging Face

What it does

  • Transcribing field and archive recordings in Karelian
  • Subtitles for Karelian-language material
  • Handling recordings that mix Karelian and Russian
  • A base for fine-tuning to your own dialect

Where it is used

Research and language projectsMuseums, archives and librariesEducation and culture

Hardware requirements

LaptopLaptop or regular PC, up to 8 GB of VRAM — smaller versions
fits
1 GPUOne GPU with 16–80 GB — mid-size versions
fits
ClusterServer with several GPUs — flagship versions
no versions

Versions

  1. wav2vec2-BERT для карельского с русским
  2. Whisper large-v3-turbo для карельского
  3. Distil-Whisper large-v2 для карельского
  4. Whisper medium, финский и карельский
  5. wav2vec2 XLS-R для карельского

How to run it

On your own serverWeights are downloaded from Hugging Face and served with vLLM (load and API) or llama.cpp (modest hardware). The model then runs in a closed network with no per-request fees.Hugging Face
How much hardware you needCalculate the VRAM for the model size, context length and number of concurrent requests.Open the hardware calculator

I can set this up end to end: pick the model size, deploy it on your server and connect it to your systems.

Frequently asked questions

Can Распознавание карельской речи (Mihaj) be used in a commercial project?

With conditions. License: no license is stated on the model cards — check with the developer. Restrictions vary — region, company revenue, attribution requirements. Have a lawyer check the terms before a commercial launch.

What hardware does Распознавание карельской речи (Mihaj) need?

At minimum: Laptop or regular PC, up to 8 GB of VRAM — smaller versions. Some versions also run on an ordinary CPU, without a GPU. You can calculate the exact VRAM for your model size and context in the hardware calculator.

Does Распознавание карельской речи (Mihaj) support Russian?

Yes, Russian is listed on the model card.

Where can I download Распознавание карельской речи (Mihaj) and what does it cost?

The Распознавание карельской речи (Mihaj) weights are open and free to download. You only pay for the hardware it runs on and for the setup. Source links are at the bottom of this page.

How I deploy it for clients

  1. SelectionI pick the model size for your task and hardware and test it on your examples.
  2. DeploymentI deploy it on your server or in a closed network and provide an API.
  3. Fine-tuningI fine-tune it on your data (LoRA) or connect a knowledge base — whichever is cheaper for the task.
  4. IntegrationI connect it to your CRM, ERP, bot, website or team chat and set up monitoring.

Similar models

Speech to textWhisperOpenAI · USACommercial use allowed

Speech recognition in 99 languages, including Russian. The de facto standard for transcribing calls and meetings. Hugging Face's faster Distil-Whisper is English only.

DetailsSpeech to textMMS (Massively Multilingual Speech)Meta AI · USANon-commercial only

Meta speech models covering more than a thousand languages. The supported list includes Bashkir, Tatar, Chuvash, Sakha, Ossetian, Chechen, Avar, Udmurt, Mari, Erzya, Moksha, Kalmyk and Karelian. Quality varies a lot by language, so test it on your own recordings.

DetailsSpeech to textРаспознавание речи для башкирского, татарского и марийского (AigizK)Aigiz Kunafin (AigizK), member of the SLONE community · RussiaCommercial use with conditions

Years of work by one enthusiast around the Bashkir Common Voice project: Whisper, wav2vec2, w2v-BERT and GigaAM fine-tuned for Bashkir, Tatar and Mari. Quality differs between models, so test them on your own recordings.

Details

Source: huggingface.co/Mihaj/wav2vec2-bert-karelian-cs-w-rus-full-ft. Data checked against the model card on 22 Sep 2026. Have a lawyer review the license before commercial launch.