E2E Networks Launches Blaze To Offer Pay-Per-Token AI Inference Services

Blaze addresses rising demand from Indian AI companies transitioning from pilots to production deployments. The platform provides cost-efficient inference with models including NVIDIA Nemotron, Sarvam, DeepSeek, Qwen, GLM and Gemma. E2E Networks' infrastructure includes approximately 14.6 MW of capacity, with a 1,024-GPU NVIDIA Blackwell cluster operational since May and 5.4 MW of additional Blackwell capacity. The service targets inference-heavy applications requiring lower latency and improved token efficiency.

Summary and key points written by AI from the headline and a short excerpt, so they can be incomplete or wrong. Analytics India Magazine did the reporting; read the original for the full story.