The State of Model Routing — NVIDIA, Cognition, OpenRouter
Walden · Carter · Tanay · Alex Atallah · Nav
AI Engineer World's Fair 2026 · 48:17
← All organizationsopenrouter.ai ↗
AI model gateway and routing platform
OpenRouter
OpenRouter provides an AI model gateway and marketplace that lets developers and businesses access hundreds of models through one interface. It supports text, image, audio, speech, transcription, embedding, and video workloads, with provider failover and cost and latency optimization. Teams can manage workspaces, spending, guardrails, and zero-data-retention policies while tracking usage across models. Users include NVIDIA, Zoom, and Lovable.
Founded in 2023 by Alex Atallah and Louis Vichy, OpenRouter is led by Atallah as CEO. Its engineering addresses differences between providers serving the same model. Auto Exacto reevaluates providers roughly every five minutes using throughput, tool-call telemetry, and controlled benchmarks. It moves statistical quality outliers behind other providers while preserving price, latency, and user preferences within each tier, making provider selection responsive to measured performance.
In August 2026, OpenRouter reported processing more than 10 trillion tokens daily for a community of over 10 million developers and companies. It raised a $113 million Series B led by CapitalG in May 2026. Stripe announced an acquisition agreement on August 19, 2026; the transaction remained subject to customary closing conditions. OpenRouter said its name, product, and roadmap would continue.
3 talks
Newest first
fun stories from building OpenRouter and where all this is going
AI Engineer World's Fair 2025 · 18:47
The Next Unicorns: 7 Top AI startups from the HF0 Residency
Diego Rodriguez · Eugene · Jonas Bauer · Shijia Liao · David Vorick · Alex Atallah
AI Engineer World's Fair 2025 · 22:16
1 speaker at AIE
Affiliations reflect their AIE appearances, not necessarily current employment.
Start here
- fun stories from building OpenRouter and where all this is going
Start here to learn how open-weight models and Stanford Alpaca challenged Atallah's assumptions about a winner-take-all inference market.
Alex AtallahAI Engineer World's Fair 2025
- The State of Model Routing — NVIDIA, Cognition, OpenRouter
Turn to the panel for discussion of frontier versus smaller models, prompt tuning, and collaborative agents.
Walden · Carter · Tanay · Alex Atallah · NavAI Engineer World's Fair 2026
Messages from the stage
Practical marketplace features
Atallah discusses provider fallback, inference plugins, MCP integration, model discovery, and prompt observability. Reducing vendor lock-in is another topic in the talk.
Application behavior and inference infrastructure
The joint routing discussion covers context compaction, evaluation and agent behavior, alongside KV-cache-aware routing and GPU cache offloading.
Affiliations reflect each recorded session, not necessarily current employment. The routing session is a joint discussion with participants from NVIDIA, Cognition, and OpenRouter. The HF0 showcase summary flags discrepancies in presenter listings and uncertainty about the conference-session match.