Perplexity Orchestrator: One API for Frontier and Open Models

A glowing central hub sending streams of light to many glass orbs representing different AI models
AI-generated illustration Anyone building with AI in 2026 faces the same quiet headache: new models arrive every few weeks, each from a different provider with its own account, contract, key and invoice. On November 17, 2026, Perplexity launched an answer: the Orchestrator API, which puts frontier and open-weight models behind a single endpoint with unified billing and automatic failover. This article draws on Perplexity's launch post. The performance statements below are the company's o...
More

Alibaba Qwen4-Max: 2.4T Parameters, Only 71B Active per Token

Glowing orange cluster of lit nodes within a vast dark blue neural network above a circuit-like landscape
AI-generated illustration Alibaba has released Qwen4-Max, and the pitch is less about being the biggest model and more about doing more work per token. Announced on November 16, 2026, the new flagship of the Qwen family has 2.4 trillion total parameters, native vision and a one-million-token context window, yet it activates only 71 billion parameters for each token it processes. What Alibaba released Per Alibaba Cloud's press release, Qwen4-Max is the most capable model in the Qwen fa...
More

MIT’s FLEX Lets Robots Read Between the Lines of Commands

Tell a robot "don't spill on the laptop" and it will dodge the laptop. Ask a person the same thing and they quietly understand that phones, tablets and anything with a battery are also off limits. On November 9, 2026, MIT News described a system that tries to close that gap: FLEX, a pipeline from MIT and New York University researchers that lets robots read meaning beyond the literal words of a correction. The work is slated for presentation at the Conference on Robot Learning this month. I...
More

Mistral Large 4 Goes Open-Weight at 1 Trillion Parameters

What was releasedOn October 29, Mistral AI released the weights of Mistral Large 4, nicknamed Le Chonk, three weeks after a public preview. According to Hugging Face's launch post, it is a multimodal, multilingual mixture-of-experts model with about 1 trillion total parameters, about 49 billion active per token, and a 1 million token context window. It takes text and images in and produces text.Two official variants are available, FP8 and NVFP4, plus community GGUF builds. The Next Web reports ...
More

GEOBEE!: The Intelligent Tablet-Based Game For Children with Learning Disabilities

GeoBee!, a new intelligent, accessible, and affordable tablet game, can provide training for skills for children with learning disabilities. Today, there is a tremendous need for vision support in school and in rehabilitation. Children and adults with developmental and language disorders are constantly looking for tools to help boost their performance. GeoBee! is the flagship game from VisionMechanic, Inc., a new tech technology company led by Dr. Charles Boulet who works on vision and motor ...
More