Cohere releases tiny-aya-global, multilingual text model covering 100+ languages
Cohere Labs has released tiny-aya-global, a lightweight text generation model trained to support conversational tasks across 100+ languages. The model is available on Hugging Face under a CC-BY-NC-4.0 license and builds on the tiny-aya-base architecture.
Cohere Labs has released tiny-aya-global, a lightweight multilingual text generation model designed for conversational tasks across 100+ languages.
Model Details
The tiny-aya-global model is available on Hugging Face and supports text generation and conversational applications. It's built as a fine-tuned variant of the tiny-aya-base model, optimized for multilingual performance across a broad language spectrum.
The model covers extensive language support spanning:
- Major European languages: English, Dutch, French, Italian, Portuguese, Romanian, Spanish, Czech, Polish, Ukrainian, Russian, Greek, German, Danish, Swedish, Norwegian, Catalan, Galician, Welsh, Irish, Basque, Croatian, Latvian, Lithuanian, Slovak, Slovenian, Estonian, Finnish, Hungarian, Serbian, Bulgarian
- Middle Eastern and South Asian languages: Arabic, Farsi, Urdu, Turkish, Maltese, Hebrew, Hindi, Marathi, Bengali, Gujarati, Punjabi, Tamil, Telugu, Nepali
- Southeast Asian languages: Tagalog, Malay, Indonesian, Javanese, Khmer, Thai, Lao, Burmese
- East Asian languages: Chinese, Japanese, Korean
- African languages: Amharic, Hausa, Igbo, Malagasy, Shona, Swahili, Wolof, Xhosa, Yoruba, Zulu
Availability and Licensing
The model is distributed under the CC-BY-NC-4.0 license and can be accessed directly from Hugging Face. As of release, the model has generated 1,204 downloads and received 56 likes on the platform, indicating early adoption among developers building multilingual applications.
The model uses the transformers library and safetensors format for compatibility with standard ML tooling.
What This Means
Tiny-aya-global fills a gap in accessible multilingual models by providing a lightweight alternative for organizations needing broad language coverage without deploying massive foundation models. The focus on lower-resource languages alongside major ones suggests Cohere's intent to extend conversational AI capabilities beyond English-dominant markets. For teams building applications in non-English regions, the model's breadth of language support reduces the need to manage separate language-specific models or expensive API calls to large closed models.
Related Articles
Inception launches Mercury 3 diffusion LLM: 1M context, 1,068 tokens/sec, $0.20/$0.70 per 1M tokens
Inception released Mercury 3, a diffusion-based LLM with a 1M-token context window that it says runs at 1,068 tokens per second on NVIDIA GPUs. Pricing is $0.20 per 1M input and $0.70 per 1M output tokens, with a 50% launch discount. Inception claims quality on par with or better than GPT-6 Luna and Claude Haiku 5.5 at 4–8x the speed.
MiniMax M3.1 Flash scores 42 on AA Intelligence Index at $0.15/$0.60 per 1M tokens, open weights Oct 23
MiniMax has released M3.1 Flash, a reasoning model that scores 42 on the Artificial Analysis Intelligence Index, up from 29 for M3. It costs $0.15 per 1M input tokens and $0.60 per 1M output tokens, and open weights are planned for October 23.
Kling AI releases 4.0: 30-second video, stereo audio, 10+ language lip sync, plus a 720p Flash tier
Kling AI has released Kling 4.0, which generates up to 30 seconds of video in one pass, adds two-channel stereo audio, and supports lip sync in more than ten languages. A cheaper Kling 4.0 Flash variant is limited to 720p and 20 seconds. API pricing has not been announced.
Mistral releases Timestral-209M, a time series model claiming top zero-shot scores on three benchmarks
Mistral AI has released Timestral-209M, the largest of four time series foundation models (5M, 28M, 123M, 209M parameters). According to the model card, it leads GIFT-Eval, TIME and fev-bench in zero-shot forecasting, with a context of up to 32,768 time steps and 21-quantile probabilistic output.
Comments
Loading...