Leading innovator in AI technology, Reka, just revealed Reka Core, their most recent creation. This innovative multimodal language model performs remarkably well across a range of assessment parameters. It is placed on par with top industry models like GPT-4V and Claude-3 Opus. Core offers outstanding value in terms of total cost of ownership, having been effectively trained from scratch over a period of several months on hundreds of GPUs.

What is Reka Core?

Reka Core is the latest flagship product from Reka, a multimodal language model. It stands out for its ability to process different types of data, including images, videos, audio, and text. Trained from scratch on thousands of GPUs over several months, Reka Core boasts competitive performance comparable to leading models in the industry. It offers a vast 128K context window, exceptional reasoning abilities, top-tier coding capabilities, and multilingual proficiency. Reka Core’s deployment flexibility, available via API, on-premises, or on-device, ensures it can meet diverse customer needs.

You can access Reka Core here.

Features and Capabilities

• Multimodal Understanding: Core can interpret and generate content from text, images, and audio inputs, offering a more complete understanding of user queries or prompts.
• Contextual Awareness: Leveraging advanced machine learning algorithms, Core excels at understanding context, allowing for more accurate and nuanced responses.
• Enhanced Creativity: With its multimodal capabilities, Core can generate diverse outputs, from detailed descriptions of images to contextual responses in natural language.
• Real-World Applications: The applications of Core are vast, from aiding in content creation to improving accessibility for individuals with visual or auditory impairments.

Impact on Industries

Reka Core has sent ripples of excitement through various industries:

• Content Creation: Core’s ability to assist in generating ideas, writing and creating multimedia content benefits writers, journalists, and content creators.
• E-Commerce: Core enhances product descriptions with images and audio, providing online retailers a more engaging shopping experience.
• Education: Core can revolutionize how students learn, offering interactive and immersive educational materials tailored to individual learning styles.
• Accessibility: By providing detailed descriptions of visual and auditory content, Core contributes to making digital information more accessible to everyone.

Partnerships

Reka’s partnership roster includes industry giants like Snowflake, Oracle, and AI Singapore. These collaborations enable broader customer use cases in industries such as e-commerce, healthcare, and robotics.

• Baris Gultekin, Head of AI at Snowflake, commented, “We are excited to bring Reka Core’s impressive industry-leading performance to customers through Snowflake Cortex.”
• Greg Pavlik, Senior Vice President at Oracle Cloud Infrastructure, added, “Reka’s dynamic multimodal and multilingual models allow enterprises to unlock more value from their data.”
• Leslie Teo, Senior Director of AI Products at AI Singapore, expressed enthusiasm about Reka’s models, highlighting their role in developing highly capable open-source models for Southeast Asia.

Our Say

Reka Technologies sees Core as the first step to advanced AI systems harmonizing with daily life. Core’s release marks a significant advancement in AI, combining language, vision, and audio processing into one model. Watch this space for updates as Reka uses Core to push the limits of artificial intelligence innovation. With its release, multimodal language models take a significant step forward. It offers hope for a time when artificial intelligence will fully comprehend human communication in all of its forms.

Follow us on Google News to stay updated with the latest innovations in the world of AI, Data Science, & GenAI.