Vera 1.6 is a frontier multimodal AI model designed for agentic systems and reasoning tasks. Built by Cortex Research, Vera 1.6 introduces a hybrid architecture with Sparse Mixture-of-Experts, enabling high reasoning performance while maintaining efficient inference at scale. The model supports text, image, and video inputs, processes extremely long documents with a 1M-token context window, and operates across 201+ languages, making it suitable for enterprise workflows and advanced AI agents.

Figr AI: UX Agent for Product Teams — Learns your product. Thinks through UX

Learns your product. Thinks through UX

Hi Product Hunt 👋 Today we're launching Vera 1.6, our newest multimodal AI model designed specifically for agentic systems and long-context reasoning. Most AI models today are built primarily for chat. But when we started building AI agents, we realised something important: real workflows require models that can reason across large documents, use tools, analyse data, and execute multi-step tasks reliably. That’s what pushed us to develop the Vera model series. Vera 1.6 introduces: • a 1M token context window for long-horizon reasoning • multimodal inputs (text, images, video) • support for 201+ languages • a hybrid architecture combining Gated Delta Networks and Sparse Mixture-of-Experts It’s designed as the intelligence layer behind the Vera Platform, where AI agents can research, analyse data, write reports, generate presentations, assist with coding, and automate complex workflows. We built Vera to enable organisations to deploy advanced AI while maintaining stronger control over infrastructure and sensitive data. We’d love your feedback, questions, and thoughts on the model! Thanks for checking it out 🙌