100+ models available through the Zima encrypted gateway. Every request end-to-end encrypted in hardware. Provider costs passed through at-rate.
Multimodal flagship with vision, code, and advanced reasoning.
Fast and affordable multimodal model for lightweight tasks.
Advanced reasoning model for complex multi-step problems.
Cost-effective reasoning model balancing speed and depth.
Most capable Claude model for complex analysis and extended tasks.
Balanced intelligence and speed for production workloads.
Fastest Claude for high-throughput, cost-sensitive pipelines.
Million-token context with native multimodal understanding.
Ultra-fast inference with million-token context at minimal cost.
Previous-gen Flash model, optimized for speed and throughput.
Open-weight model with strong code and instruction-following.
Largest open-weight model rivaling closed-source flagships.
Lightweight open model for simple tasks and prototyping.
Mistral's most capable model for complex enterprise workloads.
Efficient model for classification, routing, and simple tasks.
Purpose-built for code generation with 256K context window.
Enterprise RAG and tool-use specialist with grounded generation.
Balanced model for conversational and retrieval-augmented tasks.
State-of-the-art embedding model for search and classification.
xAI's frontier model with deep reasoning and real-time knowledge.
Compact Grok variant for fast, cost-effective inference.
Provider costs passed through at-rate. Zero markup. Pricing reflects per-million-token rates.