Just saw the announcement and got the email myself. This is huge, and frankly, a bit sooner than I expected given the limited model availability so far. As someone who's been tracking this from a procurement and vendor lock-in perspective, my immediate reaction is a mix of excitement and a flood of operational questions.
I'm absolutely considering signing up for the waitlist, but I never do that before doing my homework. The announcement blog post and documentation have some gaps that I think are critical for anyone planning to evaluate this for a real project. I'm hoping some of you who are also looking at this can share your thoughts or any insider scoops you might have.
My primary concerns before I even put my company's name on that waitlist:
* **Pricing Model & TCO:** They've announced it will be priced per second. That's a new one for me in the GenAI video space. How does that actually translate? Per second of generated video? Per second of compute time? What's the minimum chargeable duration? Most importantly, what's the estimated cost for a 10-second clip at 1080p versus a minute-long storyboard at 720p? Without a public pricing page or calculator, it's hard to gauge the total cost of ownership for a pilot project.
* **Rate Limits & Quotas:** Early access APIs often come with harsh limits. Will there be a monthly credit cap? What are the initial requests-per-minute (RPM) or tokens-per-minute (TPM) limits? This drastically affects implementation timelines if we can only generate a handful of clips per day during testing.
* **Contractual & Data Governance:** The terms of service for the API will be paramount. How is training data handled? If we generate proprietary storyboards for a client, does that input or output data get used for model improvement? What's the data retention policy on OpenAI's side? This is a major point for negotiation in any enterprise contract.
* **Output Technical Specifications:** We know it can generate variable durations, but what are the exact export options? Codec, container format, bitrate control? Can we specify a target file size? This matters for integration into existing video pipelines and storage cost projections.
* **Exit Strategy & Portability:** This is a big one. If we build a workflow around Sora's API and then need to switch providers (due to cost, performance, or policy changes), how portable is the prompt engineering? Are the prompts likely to be highly Sora-specific, or is there a chance they could be adapted to a competitor's model with similar results? I hate the idea of complete vendor lock-in.
Is anyone else going through this same mental checklist? Have you found answers to any of these points that I've missed? I'm particularly interested if there are any early whispers about the commercial terms—sometimes that info leaks in communities like this before the formal sales cycle begins.
Also, for those signing up: are you doing it as an individual developer for experimentation, or are you already getting internal approval to evaluate it for a business unit? The compliance hurdles for the latter are significant, so I'm curious how others are navigating that.