Commercial training data from specialist providers. Listings show what each source advertises; licensing is confirmed with the provider before any agreement.
Arabic Spontaneous Dialogue Smartphone speech dataset.
Specifications: ‣ Volume: 100,000+ hours ‣ Format: 8kHz, mono channel (WAV audio + JSON transcript) ‣ Environment: Studio, podcast-quality recording ‣ Recording content:…
Arabic(Saudi Arabia) Real-world Casual Conversation and Monologue speech dataset, mirrors real-world interactions.
This dataset contains human-authored communication samples in Arabic (Modern Standard Arabic), curated across common real-world interaction formats such as Email –…
This dataset contains human-authored communication samples in Arabic (Modern Standard Arabic), curated across common real-world interaction formats such as Notes –…
This dataset contains human-authored communication samples in Arabic (Modern Standard Arabic), curated across common real-world interaction formats such as SMS –…
This dataset contains human-authored communication samples in Egyptian Arabic, curated across common real-world interaction formats such as SMS – Short-form…
This dataset contains human-authored communication samples in Gulf Arabic, curated across common real-world interaction formats such as SMS – Short-form conversational…
This dataset contains human-authored communication samples in Hejazi Arabic, curated across common real-world interaction formats such as SMS – Short-form conversational…
This dataset contains human-authored communication samples in Iraqi Arabic, curated across common real-world interaction formats such as SMS – Short-form conversational…
This dataset contains human-authored communication samples in Levantine Arabic, curated across common real-world interaction formats such as SMS – Short-form…
This dataset contains human-authored communication samples in Najdi Arabic, curated across common real-world interaction formats such as SMS – Short-form conversational…
A collection of simulated conversations between two people (like whatsapp conversations) on a variety of topics.
This dataset consists of text-based records capturing human resources interview scenarios in the United Arab Emirates, framed for retail-focused analysis.
A multilingual multi-turn chat text dataset built directly by professional annotators, including speaker information, proper nouns, chat slang, typo tags, and natural…
An Arabic general-domain multi-turn chat text dataset built by professional native-language annotators for daily life and general conversation domains.