Catalogue · ಕನ್ನಡ
Ready-made Kannada speech datasets
Corpora already recorded in Kannada and available under licence, plus what to check before buying existing data instead of commissioning a build.
- Language
- Kannada (kn-IN)
- Licence
- Perpetual, commercial, model-training permitted
- Availability
- Subject to consent scope; confirmed on enquiry
- Lead time
- Days, not weeks
Available corpus types
| Corpus | Typical band | Annotation | Best for |
|---|---|---|---|
| Kannada scripted speech | 100-500 hours | Verbatim + timestamps | Prompt-read speech from a phonetically balanced script, the baseline corpus type for ASR and TTS training. |
| Kannada spontaneous speech | 100-500 hours | Verbatim + timestamps | Unscripted monologue on prompted topics, which carries the disfluencies and prosody that scripted data never produces. |
| Kannada conversational speech | 100-500 hours | Verbatim + speaker turns + timestamps | Two-party conversation recorded on separate channels, with overlap and turn-taking preserved. |
| Kannada telephony speech | 50-300 hours | Verbatim + timestamps | Narrowband call audio captured over a telephony path, matching what a deployed contact-centre model actually receives. |
Availability is confirmed per enquiry. Some material can only be licensed within the consent scope the speakers originally agreed to, and we will not stretch that to close a sale.
Ready-made versus commissioned
| Ready-made | Commissioned build | |
|---|---|---|
| Lead time | Days | 3-16 weeks |
| Cost per hour | Lower | Higher |
| Cohort control | Fixed by the original spec | Yours |
| Dialect quotas | As recorded | Designed |
| Exclusivity | Non-exclusive | Exclusive if you want it |
| Best for | Baselines, prototypes, augmentation | Production models |
What to check before licensing existing data
- Consent scope: does it permit commercial model training, and by a third party?
- Dialect distribution: a corpus that is 90% one urban variety will not generalise across Karnataka, parts of Maharashtra, Tamil Nadu and Andhra Pradesh
- Recording conditions: studio-only data underperforms badly on noisy deployments
- Transcription convention: Northern lexical items replaced with standard equivalents will otherwise surface as tokenisation noise
- Overlap: check whether the corpus already sits in the public sets you trained on
Where the public corpora fall short
Mysuru/Bengaluru standard dominates. North Karnataka (Dharwad, Kalaburagi) and coastal Mangaluru speech are barely represented in any public corpus.
This is usually the reason buyers move from licensing to commissioning: the free and cheap material covers the easy half of the language.
What a usable Kannada corpus has to cover
- Dialect spread across Bangalore urban, Mysuru (standard literary), Dharwad / North Karnataka, Mangaluru coastal rather than a single prestige variety
- The phonetic contrasts that Kannada models actually get wrong: North Karnataka speech has markedly different intonation and lexicon from Mysuru standard
- Code-mixed speech transcribed rather than discarded — Bengaluru is a migration city: Kannada speech there is mixed with English, Hindi, Tamil and Telugu. Native-only Kannada cohorts recruited in Bengaluru are hard to fill without screening for years of residence.
- Recording geography covering Bengaluru, Mysuru, Hubballi-Dharwad, Mangaluru, since dialect follows location
- Utterance types beyond read prompts: Dialect-tagged spontaneous conversation across five zones, Read prompts balanced for retroflex and geminate consonants, Ride-hailing, delivery and fintech domain command sets
Screen Bengaluru participants for native fluency and years of Karnataka residence; otherwise the cohort drifts towards second-language Kannada.
Frequently asked
Can I buy an existing Kannada dataset today?
Where a corpus exists within an appropriate consent and licence scope, yes — usually within days. We confirm availability against your intended use before quoting.
Is the licence exclusive?
Ready-made corpora are licensed non-exclusively. Exclusivity is available on commissioned builds.
Can we augment a ready-made corpus with new recordings?
That is the most common pattern: licence the base, then commission the dialects, conditions or domain vocabulary it lacks.
Do you provide a data sheet?
Yes — speaker counts, demographic distribution, condition mix, annotation convention and known limitations, before you commit.
Related pages
Check Kannada availability
Tell us the hours, corpus type and intended use. We confirm what is licensable now and what needs recording.