40+ World-Class Models

NVIDIA's 550B MoE reasoning model — 1M context, 55B active params, open-source, free tier via OpenRouter.

Alibaba's massive-scale flagship — native Chinese mastery with unbeatable cost-performance.

OpenAI's flagship multimodal model — blazing fast inference, top-tier reasoning and vision.

128K extended context window — ideal for complex reasoning and enterprise-grade code generation.

Anthropic's latest speed-intelligence hybrid — perfect balance of blazing speed and elite coding.

Heavy-duty reasoning engine designed for the toughest logic problems and deep analysis.

The open-source coding champion — top-tier code generation at a fraction of proprietary costs.

Google's frontier model with a massive 1M+ token context — ideal for large-scale document analysis.

Meta's powerful open-source LLM — versatile, reliable, and optimized for general chat and instruction following.

NVIDIA 120B MoE reasoning model — 12B active, fast inference, free on OpenRouter.

NVIDIA 30B multimodal MoE — 3B active, vision + reasoning, free on OpenRouter.

NVIDIA 30B lightweight MoE — 3B active, ultra-fast, free on OpenRouter.

NVIDIA's smallest production model — 9B params, edge-deployable, free on OpenRouter.

NVIDIA's dedicated content safety model — toxicity detection, PII masking, compliance screening, free on OpenRouter.