Imagine every extra millisecond of latency stealing tokens — and therefore intelligence Every millisecond of AI latency steals intelligence from your model and delight from your user experience. In this high-energy session, Cerebras’ VP of Product Management pulls back the curtain on our wafer-scale inference platform that turns that millisecond tax into pure headroom for smarter model reasoning and higher user engagement. We’ll trace the industry’s pivot from “bigger pre-training” to “smarter run-time,” show why data scarcity and spiraling training budgets make inference the new competitive front, and reveal how models like Qwen3 and Deepseek gain IQ-like leaps by "thinking for longer". Expect live numbers, behind-the-scenes engineering stories, and a first look at the architectural tricks that let us stream 10× more tokens while cutting power in half. If you care about building agents that think in real time, not coffee-break time, this is your roadmap to the fastest inference on Earth—and the dawn of AI’s next era.

The age of specialized, limited AI agents is over.
The era of an all-in-one General AI Agent that actually executes has begun.
‍
Meet Super Ninja — the General AI agent that doesn't just handle one piece of your project; it completes the entire workflow from start to finish.
While other agents get stuck with token limits or require constant hand-holding, Super Ninja runs extensive data analysis, codes and validates full applications, does comprehensive research, builds websites, and delivers high-quality results in your preferred format.
Powered by our proprietary models on Cerebras' wafer-scale architecture, it operates like having a team of experts working 24/7.
‍
This isn't just another AI tool — it's the foundation for the next generation of autonomous AI Agents and digital robots for personal & business productivity.

In this fireside chat Meta will share the latest Llama developments, including the launch of Llama 4 and Llama API preview. Meta will discuss how the Llama enables developers to build and deploy applications easily and how the Llama API is delivering faster and more efficient inference for the world in partnership with Cerebras.

Picking up where the previous deep-dive into Cerebras’ wafer-scale architecture left off, this session zooms in on four concrete, high-impact use cases unlocked by sub-millisecond inference.
Voice, Digital Twins, Code Generation, and Agents.
‍
Each use case comes with the key performance stats—latency envelopes, token budgets, and cost profiles—that turn ambitious prototypes into reliable products.
‍
Attendees will leave with practical blueprints for bringing Cerebras speed into their own applications and a clear view of where ultra-fast inference is headed next.

What if your customers could navigate your website or fill out forms on your mobile app simply by talking?
Join Andrei Papancea, CEO of NLX, for a look at a future without clicks and scrolls. He will unveil Voice+, the company's patented multimodal technology that transforms any digital property into a truly interactive conversational interface.
‍
‍Discover how Voice+ allows users to not only talk with an AI but to talk to the user interface itself — driving navigation, completing forms, and taking action using only their voice.
Powered by Cerebras's blazing-fast AI inference, these interactions are delivered with zero latency, making them as seamless and intuitive as a human conversation.
‍
Andrei will dive into real-world applications, showcasing how this revolutionary approach is creating truly hands-free digital experiences and defining the next era of human-computer interaction.

What does it take to bring cutting-edge AI to over 100 million users — and actually make it useful?
In this fireside chat, Notion’s Tian Jin and Cerebras’ Angela Yeung go behind the scenes on building practical, production-ready AI that feels magical but works at scale.
‍
They’ll explore how Notion approaches critical product decisions like when to fine-tune a model, how to evaluate quality across different models and agentic pipelines, and what “vibe working” looks like for the next generation of productivity tools.
You'll get a behind-the-scenes look at Notion’s AI stack — from how model routing frameworks help match the right model to the task, to where fast inference from Cerebras enables instant, more responsive user experiences.
‍
This conversation is for builders who care about delivering great AI experiences: real systems, real trade-offs, and building AI with an intense focus on delivering the best possible user experience at scale.

Building with Llama: The Future of AI Development

‍In this fireside chat Meta will share the latest Llama developments, including the launch of Llama 4 and Llama API preview. Meta will discuss how the Llama enables developers to build and deploy applications easily and how the Llama API is delivering faster and more efficient inference for the world in partnership with Cerebras.

Enterprises today face a critical challenge: how to easily deploy and manage a diverse set of AI models without sacrificing performance.
This talk introduces the powerful combination of IBM Model Gateway, providing a single, secure interface to access, govern, and deploy foundation models, and Cerebras, the world’s fastest inference provider.
In our live demo, you’ll see firsthand how to simplify your AI lifecycle, maintain enterprise-grade control, and serve models at breathtaking speed — all through one unified platform.

Cerebras gives you blazing-fast inference.
Docker Compose now gives you the simplest way to run agentic apps — with just a compose.yaml.
Define your open models, agents, and Cerebras API endpoints. Then spin up your entire agentic stack.
From local testing to full-scale deployment, your Cerebras-powered agents are wired and running in seconds.
No rewires. No config gymnastics. Just fast.

Agentic pipelines are becoming increasingly sophisticated, integrating components such as RAG modules (e.g., vector databases, embedding models, retrievers), as well as verifiers, rewriters, and rerankers.
Each module introduces complex configuration and hyperparameter choices, resulting in a vast design space with different trade-offs between latency, accuracy, and cost.
Syftr is a novel framework that automates this exploration of the space, using multi-objective Bayesian Optimization to surface Pareto-optimal agentic pipelines.
In this presentation, we explore how pipelines running on the Cerebras Wafer-Scale Engine perform when optimized by Syftr for low-latency use cases.

Join AI leaders from GSK and Cerebras for an inside look at how wafer-scale compute is reshaping life-sciences R&D.
In this fireside chat, they trace the journey from the first training runs — where Cerebras’ single-chip architecture slashed model-training cycles from weeks to days — to today’s blistering inference that powers new breakthroughs in seconds.
Attendees will leave with a clear picture of how the two teams co-engineered hardware, software, and scientific models to turn breakthroughs in computers into breakthroughs across the life-sciences spectrum.

Dr. Shant Ayanian's talk at the Cerebras Supernova event will highlight the Mayo Clinic's use of Cerebras' AI infrastructure to pioneer genomic-based personalized medicine.
This collaboration will focus on creating individualized treatment models by integrating genomic and clinical data, moving beyond traditional "one-size-fits-all" approaches.
The discussion will emphasize the transformative potential of AI in healthcare and the importance of partnerships in advancing medical treatments.

Let’s live long and prosper! Join us in the Supernova Lounge for the Longevity Happy Hour hosted by RAISE.

Head of Growth at Cerebras, Daniel Kim, will be sharing how higher-scale inference has been unlocking insane creativity and diverse use cases for developers and startups.
The unique design of the Cerebras wafer enables breakthroughs in AI performance, power efficiency, and scalability, and Daniel will explore real-world use cases that leverage these capabilities.
Through this talk, he aims to demonstrate the vast potential of Cerebras to accelerate AI adoption and drive innovation in industries such as healthcare, finance, and more.

In this hands-on workshop, you’ll build a real-time voice-based AI sales agent using Cerebras for fast inference and LiveKit for conversational flow.
The agent can speak, listen, and respond based on real sales context — no hallucinations.

Learn how to build a Perplexity-style AI assistant that performs deep research in under 60 seconds.
In this workshop, you’ll combine Exa's web search and Cerebras' fast inference to scrape, analyze, and synthesize content from multiple sources into structured insights.
You’ll walk away with a working AI research agent that can summarize events, extract key takeaways, and scale to any topic.