Browse every model routed through Kyara Intelligence. Pick the right one for your workload, or let us choose automatically.
x-ai/grok-4.1-fastxAI agentic tool-calling model for support and deep research. 2M context, 30K max output.
mistralai/mistral-nemoMultilingual 12B model with 128k context, built by Mistral and NVIDIA.
deepseek/deepseek-v4-flashEfficiency-optimized 284B MoE model with 1M context for fast inference.
z-ai/glm-5.1Long-horizon coding model that can work autonomously for 8+ hours.