Sora

OpenAI's video model — app closed April 2026, API sunsets Sept 2026

Video & Audio Paid Has API
Researched · Published · Reviewed
RECATOOLS Score
3.5 / 10
Capability
9
Value for money
6
Ease of use
7
ASEAN readiness
6
API quality
6
Founded
2024
HQ
San Francisco, California
Users
~9.6M lifetime app downloads; ~1M peak active users, falling under 500K by shutdown
Launched
Model: Dec 2024 (in ChatGPT) | Standalone app
Developer
OpenAI

Overview

Sora is OpenAI's text-to-video generation model, unveiled publicly in February 2024. It generates photorealistic video of up to 60 seconds from text descriptions, with a level of physical plausibility, cinematic coherence, and character consistency that significantly exceeded anything previously available. The current model, Sora 2 (released September 30, 2025), became the only version available after sora.com fully switched over in March 2026.

OpenAI shut this down in two stages: the consumer web/app was discontinued April 26, 2026 (citing roughly $1M/day operating cost against $2.1M in total lifetime revenue, a collapsed Disney licensing deal, and a profitability push ahead of an IPO), while the developer API is on a separate, later sunset scheduled for September 24, 2026 — meaning the API remains live today even though the consumer app is gone. Sora continues internally at OpenAI as a research project on "world models."

Advertisement

Pricing

Pricing shown for reference only. These figures reflect RECATOOLS research as of 11 Jul 2026 and may be out of date or incomplete. This is not financial or purchasing advice — always confirm the current price on the provider’s official website before making any decision.

Consumer app
Discontinued
The Sora web and mobile apps were shut down on 26 April 2026.
  • No longer available to new or existing users
  • Video generation moved out of the standalone app
Sora 2 Pro API (legacy)
From $0.30/sec
Higher-quality legacy API tier, also sunsetting 24 September 2026.
  • From $0.30 per second generated
  • Higher resolutions cost more per second
  • Existing keys only

Use cases

Generating cinematic b-roll for documentary and commercial film production Creating product visualisation videos from text descriptions Producing concept animations for pitch decks and creative presentations

What you can produce with Sora

  • A short photorealistic video (up to ~15s on Plus, ~25s on Pro for Sora 2; up to 20s on the original 2024 model) from a text prompt, with physics-accurate water, fire, or crowd motion
  • A social-native vertical video clip (TikTok/Instagram Reels format) with synchronized dialogue and ambient sound effects generated in one pass
  • A concept reel or pre-production storyboard animatic for pitching a film scene or ad campaign to clients
  • A 'Cameo' personal video where the user's own likeness is inserted as a character into a generated scene
  • A product visualization video animating a still image or design sketch into a short motion clip
  • A stylized animated short — ranging from cinematic live-action to surreal or cartoon aesthetics — generated from a single descriptive prompt
  • A multi-shot sequence with consistent character identity across three cuts, suitable for a brief narrative or brand story draft
Advertisement

ASEAN Perspective

Sora in Southeast Asia

ASEAN-region availability and pricing notes coming soon. Drop the editorial team a note via /contact/ if you can supply local context (Singapore/Malaysia/Indonesia/Thailand/Vietnam).

RECATOOLS Verdict

Note: the consumer Sora product was discontinued 26 April 2026; only the developer API remains, sunsetting 24 September 2026.

Sora is winding down. The consumer app and sora.com were discontinued on April 26, 2026 — about seven months after the Sora 2 launch — and the API follows on September 24, 2026, so anything built on it needs a migration plan now. On raw capability the model sat near the top of AI video: strong physics, synchronized audio, and the Cameos feature were genuinely ahead of the field. But OpenAI couldn't make the economics work (reported ~$1M/day compute against $2.1M lifetime in-app revenue) and is refocusing on coding and enterprise products. Practical guidance: don't adopt. Existing API users should evaluate Google's Veo, Runway, or Kling before the September cutoff. The technology survives inside OpenAI as world-model research and will likely resurface elsewhere, but as a tool you can actually use, its clock has months left.

Independent AI-assisted assessment by RECATOOLS.

What people say

Few products have gone from No. 1 on the App Store to shut down in seven months. The Sora 2 app launched September 30, 2025, logged around 164,000 installs in its first 48 hours, passed a million downloads inside five days, and topped the US charts. By April 26, 2026 the consumer app and web experience were gone. OpenAI announced the two-stage wind-down on March 24, 2026; the developer API gets a longer runway, sunsetting September 24, 2026, so it still answers calls today.

The economics were the killer. Reported operating costs ran around $1M per day while in-app purchases produced about $2.1M over the product's entire life. Downloads peaked above 3.3M in November 2025 and fell to roughly 1.1M by February 2026; active users dropped from near 1M to under 500,000. A collapsed Disney licensing arrangement and OpenAI's pre-IPO profitability push — a broader retreat from consumer creative tools toward coding and enterprise — sealed it.

While it lived, testers genuinely rated the physics realism, synchronized audio, and Cameos (inserting yourself into generated scenes). Recurring gripes: short clip caps, weak character consistency across shots, restrictive guardrails, a demo-vs-reality gap on complex scenes, and the API arriving only two weeks after the app. No meaningful G2 or Capterra review base ever formed — this was a consumer app, rated in app stores, and it now reads as the case study in how brutal consumer AI video economics are. Sora continues inside OpenAI as world-model research.

Summary of public user & expert reviews, compiled by RECATOOLS.

Notable facts

  • Sora's February 2024 preview demonstration shocked the film industry — directors at Sundance described the technology as 'existential' for traditional visual effects workflows.
  • The model was trained partly on licensed video content from Shutterstock, making it one of the first major AI video models with a transparent content licensing strategy.
  • Sora can generate video in the style of different camera lenses — from wide-angle drone shots to close-up handheld footage — responding to cinematography terminology in prompts.

Frequently asked questions

Is Sora free?
No. Sora requires ChatGPT Plus ($20/month) for limited access or ChatGPT Pro ($200/month) for 1080p and longer videos.
How long can Sora generate video?
Up to 20 seconds at 1080p on Pro; shorter clips on Plus.
Can I use Sora-generated videos commercially?
Yes. OpenAI permits commercial use of Sora output.
How does Sora compare to Runway?
Sora produces longer, higher-quality video with better physical consistency. Runway is more accessible, has more controls, and has broader availability.
Does Sora have an API?
Not publicly available as of mid-2025. Access is only via the ChatGPT interface.

About this listing

Researched on
Published on
Last reviewed

This entry was compiled from publicly available data including Sora's official website, press releases, documentation, and reputable third-party publications. RECATOOLS is not affiliated with Sora unless explicitly stated.

Data accuracy

Third-party AI tools update their pricing, features, availability, and policies frequently. Information here may be outdated by the time you read this — we make reasonable efforts to keep listings current, but cannot guarantee absolute accuracy.

For the latest details, please refer to Sora directly →

Spotted something out of date? Suggest an update →

Advertisement