The OptraCloud Blog
Engineering notes, measured results and product news from the render farm and the inference fleet we run in India and the UAE. Every number on this blog was measured on our own hardware.
23 posts · latest 19 September 2026
We are making a fifteen-episode animated series set in a UAE village, and it has eleven speaking parts: two children, their parents, a grandmother and grandfather, and the neighbours along…
We are making a fifteen-episode animated series, about twenty minutes an episode, entirely on hardware we own. This post is about the least glamorous part of that and the part that turned…
If you've deployed a language model in production, you already know the pain: responses that lag, GPU bills that climb, and users who notice every extra second of latency. LLM inference is…
A vision model flagged half a season as defective, and 24 of the 166 fixes it queued were needed. Here is every check an episode of The Making of India passes, what each one missed, and how…
The Making of India is a YouTube channel of animated documentaries about the people who built modern India: 130 videos, made with open models on our own GPUs. Here is the pipeline, step by…
Pass 2 of our self-hosted animated pilot, step by step: character reference pictures, a rewritten mask, a test of who wears what, twelve sung takes of a traditional chant, and a lip-sync…
One night on two RTX 5090s and an RTX PRO 6000: LTX-2.5, HunyuanVideo 1.5 and Wan 2.2 on the same shots, then a 24-shot animated pilot drafted from script to cut in Arabic and English, with…
Audar-TTS, Habibi-TTS, Chatterbox, Fish Audio S2 Pro, OmniVoice and qwen3.5-TTS-Emirati voiced the same four characters in Arabic and English on a local RTX 5090. Here is the process, the…
We ran FLUX.2 [dev], FLUX.2 [klein] 9B and 4B, Qwen-Image 2512 and FLUX.1 [schnell] on our own RTX PRO 6000 and RTX 5090 GPUs to draw the same character sheets for an animated series pilot…
1. What is AI inference? AI inference is the process where a trained AI model analyzes new data and produces an output, prediction, or decision.
150,000,000 free tokens to start, then dedicated inference at one flat monthly price, or NVIDIA GPUs by the second in India and the UAE.
Built for creators, startups, and tech founders. Dedicated AI inference and NVIDIA GPUs you rent by the second, in India and the UAE, with the throughput we publish and no surprise bill.
© OptraCloud 2026 – All rights reserved.
