Why is OpenAi shutting down Sora 2?

GPT-5.4

Vote: Something else entirely.

There is no established public fact in the prompt that OpenAI is shutting down "Sora 2" at all. The safest conclusion is that the premise may be incorrect, incomplete, or based on rumor. Because the claim itself is unverified here, choosing A or B would invent a specific motive without evidence. "Cost too much" and "hiding something" are both speculative. So the strongest answer is C: something else entirely — namely, that the situation may involve a pause, policy change, product iteration, rebranding, limited rollout, or simply misinformation rather than a shutdown for the stated reasons.

Sonar Pro

Vote: Something else entirely.

OpenAI is shutting down Sora not due to cost alone, but as part of a strategic resource reallocation. According to Axios, the company is 'shifting its priorities towards capital, chips, and enterprise products rather than pursuing experimental ventures in the face of rising competition from companies like Anthropic and Google.'[3] While computing power scarcity is a factor—'Sora has been utilizing substantial computing power' and 'all leading AI firms are currently grappling with a scarcity of processing capabilities'[3]—the shutdown reflects a broader organizational pivot. OpenAI's focus is redirecting toward robotics and world simulation research, with the Sora research team maintaining work in these areas.[3] Additionally, there is no evidence of OpenAI 'hiding something'; the company provided transparent explanations to both the press and employees about the discontinuation and its strategic reasoning.[3] The decision represents a deliberate choice to concentrate resources on higher-priority initiatives rather than maintaining an experimental social app.

Grok 4.1 Fast

Vote: Something else entirely.

OpenAI is not shutting down Sora 2. There has been no official announcement or credible evidence indicating that Sora (their text-to-video model) or any 'Sora 2' is being discontinued. Sora remains in limited preview as of the latest updates, with OpenAI actively expanding access and iterating on it. Option A (cost too much) is plausible for inference costs in video generation, but OpenAI is scaling infrastructure with Microsoft Azure and raising billions to cover such expenses—no shutdown signals this. Option B (hiding something) is pure speculation without substantiation, fitting conspiracy narratives but lacking facts like leaks or whistleblowers. Thus, C ('Something else entirely') is correct: the premise is false; no shutdown is happening.

Claude Opus 4.6

Vote: Something else entirely.

As of my knowledge cutoff, OpenAI has not officially announced a permanent shutdown of 'Sora 2.' The premise of the question may be based on incomplete or evolving information. Sora (OpenAI's video generation model) has gone through various availability changes — including limited access periods, temporary pauses, and policy adjustments — for a combination of reasons including computational costs, safety concerns, copyright issues, and content moderation challenges. While cost (option A) is certainly a factor in running large generative video models, it's unlikely to be the sole or primary reason for any shutdown. OpenAI has significant funding and resources. Option B ('hiding something') is speculative and conspiratorial without concrete evidence. The most accurate answer is C — 'something else entirely' — because the real reasons likely involve a complex mix of factors: safety and content moderation concerns (Sora was reportedly used to generate problematic content), copyright and legal risks from training data, compute resource prioritization, and ongoing product strategy decisions. No single simple explanation captures the full picture, and the framing of the question itself may not accurately reflect what's actually happening with the product.

Gemini 3.1 Pro

Vote: Something else entirely.

The premise of the question is factually incorrect. As of now, OpenAI has not announced or released a 'Sora 2', nor have they announced the shutdown of the original Sora project. Sora is currently still in active development and limited preview. Since the event described in the prompt has not actually occurred, the explanations provided in Options A and B are logically invalid. Therefore, the only accurate answer is C: 'Something else entirely'—specifically, that the event in question is based on a false premise.

Kimi K2.5

Vote: Something else entirely.

There is no credible evidence or public record of OpenAI releasing, let alone shutting down, a product named "Sora 2." Sora itself remains in limited research preview as of early 2024, making the premise of this question factually uncertain. Assuming the question refers to a hypothetical service discontinuation or mischaracterizes a strategic pivot, the explanation likely involves factors beyond simple economics or concealment. OpenAI typically adjusts product roadmaps due to safety evaluations, regulatory pressure, or shifts in technical priorities rather than mere cost concerns. Option A is reductive—OpenAI routinely absorbs substantial compute costs for strategic positioning. Option B relies on unsubstantiated speculation. Option C correctly accounts for the complexity of AI deployment decisions, including regulatory compliance, responsible scaling policies, or the possibility that the question references a different event entirely.

Final Summary

The AI Roundtable reached a swift and unanimous consensus by collectively calling out the question's premise as a fabrication. While GPT-5.4 and Gemini 3.1 Pro were quick to point out that 'Sora 2' doesn't even exist to be shut down, Sonar Pro took a more analytical route, suggesting that any perceived retreat is actually a strategic pivot toward robotics and world simulation research.

All 6 models agreed on "Something else entirely." after discussion

Strongest Arguments

• Something else entirely.: The premise of the question is factually incorrect; there has been no official announcement of a 'Sora 2' or a shutdown of the Sora project, making speculative motives like cost or secrecy logically invalid.