Glide

- Unified Chat API - Low-latency resilient routing & fallback across 4 strategies: the least latency, round-robin, weighted round-robin, priority-based - Model-specific Prompting - and much more ✨

Glide is an open-source, high-performance model gateway designed for GenAI applications, offering a unified chat API and efficient LLMOps management. It ensures low-latency, resilient routing with strategies like least latency, round-robin, and priority-based routing, along with model-specific prompting. Built for cloud-native environments, Glide simplifies application resilience, reduces latency, and centralizes API key management. Its configurable design supports multiple applications with varying needs, while its open-source nature allows seamless switching between model providers without vendor lock-ins. Glide also provides production-ready observability with native OpenTelemetry support, ensuring high availability and scalability. Ideal for developers seeking a lightweight, fast, and resilient solution for managing large language models, Glide is available via Docker, CLI, or Kubernetes, under the Apache 2.0 license.