Magahi AI Training Datasets
Magahi (मगही) spoken in Bihar (East India), written in Devanagari, is one of the languages we build proprietary, rights-cleared datasets for. It is a low-resource language: usable training data barely exists on the open web, so it has to be collected from real speakers with consent. Every Magahi dataset ships with documented consent, provenance and published quality metrics, and is available with India data residency.
Magahi datasets in the catalog
Showing 0 of 41 datasets.
No published datasets here yet
We don't have an off-the-shelf Magahi dataset published yet — but we build them to spec. Tell us what you need and we'll collect or annotate it, rights-cleared and documented.
Services for Magahi data
When the data you need doesn't exist yet, we build it — collection, annotation, alignment and evaluation, all rights-cleared and documented.
Multilingual Data Collection
Native-speaker audio, video, image and text collection across Indic, African and low-resource languages — field-grade, consented, documented.
Explore the serviceData Annotation
Speech, NLP, CV and multimodal annotation at IAA ≥ 0.85 with two-pass QA — built for foundation-model SLAs.
Explore the serviceSFT Gold-Standard Data
High-quality prompt/response curation for supervised fine-tuning — vetted by linguistic SMEs, not crowd-sourced noise.
Explore the serviceCultural & Cross-Lingual Evaluation
Evaluation for honorifics, code-mix, idioms, caste-safety and pragmatic correctness — beyond translated MMLU.
Explore the serviceCommon questions
- Is the Magahi data rights-cleared and safe to train on?
Yes. Every Magahi record is created or sourced under written contributor agreements granting commercial reuse, with a consent reference and authorship log — never scraped.
- Can you collect more Magahi data to spec?
Yes. When the Magahi data you need doesn't exist, we collect and annotate it to your specification across speech, text and multimodal modalities, with an India-residency option.
- How is Magahi data quality measured?
Each dataset ships with a data card: inter-annotator agreement, QA pass rate, and (for speech) word error rate, with two-pass QA and adjudication.
Need Magahi data your model can train on?
License what's in the catalog, or tell us exactly what you need — we build proprietary, rights-cleared datasets to spec, with India data residency.