🔥The 9th Round of Easy Loan, Earn $40 Reward is in progress❗️ ⏰ Promotion Period: January 15th - Feburary 15th, 2025 👉 Register now and check more details at gate.io/campaigns/358

Aya Expanse from Cohere For AI can chat, listen, speak, and generate image in 23 world languages🔥

Aya-Expanse is a really impressive model if you need a highly performant pre-trained LLM (or speech) in one of the below 32 languages. Congratulation Cohere For AI and cohere Easy to combine with speech and images as well (just play with the HF Space below) Languages: Arabic,

Have you had a chance to check out Aya Expanse on Hugging Face Space yet? 👀

So happy to be a part of this! 🌎🌍🌏

A deeper look into MMLU showing that a large proportion of the Qs are culturally biased. The careful annotation of Culture-Specific and Culture-Agnostic can be used to better test the LLMs’ multicultural capabilities. Happy to have contributed to this!

We will be presenting this work next week at #NeurIPS2024 as a maintrack paper on Tuesday. Congrats to all the authors, Meriem Edward Kim Beyza Ermiş Marzieh Fadaee 🔥. Most of us will be at NeurIPS next week to catch up!

Aya Expanse technical report is out! Was a pleasure working on Aya Expanse with the most special team and the amazing Aya community! ❣️ Try out the model here: huggingface.co/spaces/CohereF… And learn more about how we did it here 👇

Cohere For AI is proud to support the open weights release of the new Command R7B model. This is part of our continued effort to make breakthroughs accessible to the research community.

Excited to share we are hiring for a new type of scholar role. Algorithm interface scholars will re-imagine how humans interact with algorithms. These scholars will lead the frontier in UI-algorithm co-design. 🔥 Join us at Cohere For AI cohere ✨ jobs.ashbyhq.com/cohere/dc7da5f…

Introducing ✨ Aya Vision ✨ - an open-weights model to connect our world through language and vision Aya Vision adds breakthrough multimodal capabilities to our state-of-the-art multilingual 8B and 32B models. 🌿

Super proud to introduce Aya Vision 🌿👁️— our 8B and 32B multilingual vision-language models! Aya Vision models excels in 23 languages, bringing true multilinguality to multimodal AI. Making AI more inclusive, accessible, and globally impactful🌍.

"We do not describe the world we see, we see the world we can describe." René Descartes Very proud to release Aya Vision 🌿 today, which expands the worlds AI can see. We pushed very hard to build something efficient, accessible and global. This is an important step forward.

Today cohere is very excited to introduce Command A, our new model succeeding Command R+. Command A is an open-weights 111B parameter model with a 256k context window focused on delivering great performance across agentic, multilingual, and coding usecases. 🧵

Very excited to release Kaleidoscope—a multilingual, multimodal evaluation set for VLMs, built as part of our open-science initiative! 🌍 18 languages (high-, mid-, low-) 📚 21k questions (55% require image understanding) 🧪 STEM, social science, reasoning, and practical skills

There's a new paper circulating looking in detail at LMArena leaderboard: "The Leaderboard Illusion" arxiv.org/abs/2504.20879 I first became a bit suspicious when at one point a while back, a Gemini model scored #1 way above the second best, but when I tried to switch for a few

1/ Science is only as strong as the benchmarks it relies on. So how fair—and scientifically rigorous—is today’s most widely used evaluation benchmark? We took a deep dive into Chatbot Arena to find out. 🧵

When a benchmark becomes the standard, it naturally becomes a target. The real challenge is keeping it robust — which is exactly what we explore in our latest work: 📄 arxiv.org/pdf/2504.20879

Excited to share our technical report on Aya Vision 🥳 We dive deep into what we learned, including: 🧱a synthetic multilingual data framework 🏗️architecture, training recipe and insights 🧲cross-modal model merging 📊full eval on open-ended, multilingual, generative tasks

Our new Aya Vision report is out! 🚀 Big congratulations to Saurabh Dash, Yiyang Nan, and the fantastic team at Cohere Labs cohere! 🎉 📄 Technical report: arxiv.org/abs/2505.08751 🛠️ Models and datasets: huggingface.co/collections/Co…

Super thrilled to share GMMLU is accepted to #ACL2025 main conference 🎉 It was also recently recognised by Stanford HAI as one of the significant AI releases of 2024 🚀 I had a blast collaborating on this closely with Beyza Ermiş and all our collaborators! Huge congrats!💙