E2E Networks Launches 'Blaze' AI Inference Service

E2E Networks Limited has launched 'Blaze', a new pay-per-token inference service for open AI models. Built on NVIDIA DSX OS and GPU infrastructure, Blaze offers a single OpenAI-compatible API for leading open models like NVIDIA Nemotron, Sarvam, and others. The service is designed for inference-heavy products such as coding agents, voice applications, and document AI, catering to both domestic and international markets. Customers pay only for the tokens they use, addressing critical needs for inference cost, latency, and data sovereignty in India's growing AI ecosystem. Blaze integrates with E2E's AI platform, TIR, providing a path from training to production inference.