Aya Vision: Multilingual Multimodal AI Advancements

Research

May 14, 2025 Aya Vision: Multilingual Multimodal AI Advancements To address the challenges of building multimodal language models, we introduce novel techniques spanning both data & modeling, resulting in Aya Vision 8 B and 32 B models.

Visit Site

Research Cohere

ResearchSEACrowd: A Multilingual Multimodal Data Hub and Benchmark Suite for Southeast Asian LanguagesCohere ResearchBigScience: A Case Study in the Social Construction of a Multilingual Large Language ModelCohere ResearchOne Tokenizer To Rule Them All: Emergent Language Plasticity via Multilingual TokenizersCohere ResearchPolicy Primer - Efficient AICohere ResearchUnderstanding and Mitigating Language Confusion in LLMsCohere ResearchFAIR-Ensemble: When Fairness Naturally Emerges From Deep EnsemblingCohere BlogsMind the Gap: The Chasm Between AI Fiction and Fact - Deepgram Blog ⚡️Deepgram BlogsHow Hunch supercharged AI workflows with Modal SandboxesModal Blogsbunny.net expands into Nairobi, KenyaBunny BlogsHow Walsall Council Delivers Improved Resident Experiences Through Multilingual AI Videos in Over 50Synthesia NewsCirrascale Cloud Services® and Cerebras Systems Announce Availability of Cerebras Cloud @ CirrascaleCerebras BlogsGemma 4 on Cerebras—The Fastest Inference is Now MultimodalCerebras BlogsAI-generated Awareness Campaigns: Global ImpactHeygen LearnFlux Multilingual Technical Deep Dive: Multilingual Speech-to-Text Without the Routing MessDeepgram BlogsHow Cision uses AI video to scale multilingual support and improve client experienceSynthesia BlogsKimi K3 by Moonshot now available on ModalModal ResourcesApple Safari - Glossary | MDNDeveloper Mozilla

BlogsBunnyCDN is evolving! Introducing bunny.net!Bunny

BlogsCreate Comic Books From Videos with yack! - Deepgram Blog ⚡️Deepgram BlogsVideo Localisation & Multilingual VideosHeygen