Sora is OpenAI's text-to-video AI — the product that turned generative video from a research curiosity into a consumer category. Announced in February 2024 with a set of demos that went viral, shipped to ChatGPT subscribers in December 2024, and relaunched as Sora 2 with audio and Cameo in September 2025, Sora is the highest-profile product in the generative video space.
This page is the living reference to Sora — the model, the consumer app, pricing, capabilities, controversies, and every story we publish tagged Sora. We update it whenever OpenAI ships a meaningful change.
As of April 2026, Sora 2 is the current model, with commercial use rights bundled into ChatGPT Pro and enterprise access via the Sora API.
Latest news about Sora
Every story we've published tagged Sora. Refreshed automatically every time the newsroom publishes.
- Aug 18, 2026Robin Williams' children reactivate his Instagram to counter AI deepfakesZelda, Zak, and Cody Williams say the account will host authentic clips of their father as AI-generated videos of the late actor proliferate.
- Jun 13, 2026Tribeca 2026 shows AI filmmaking needs custom models, not vanilla promptsGoogle DeepMind and OpenAI brought AI shorts to Tribeca. The standout used custom Veo and Imagen builds trained on hand-painted Pixar concept art.
- Jun 7, 2026OpenAI plans ChatGPT super app revamp, declares 'Chat is dead'The company is folding Codex and agents into ChatGPT to push free users toward paid products ahead of an IPO.
Timeline of major releases
- Feb 15, 2024Sora announcedOpenAI previews Sora with a set of research demos that go viral. The model shows multi-shot coherence and physical plausibility far beyond prior text-to-video.
- Dec 9, 2024Sora 1 shipsOpenAI launches Sora as a consumer product bundled with ChatGPT Plus ($20/mo) and Pro ($200/mo). Initial version produces up to 20 seconds of 1080p video.
- Mar 10, 2025Sora StoryboardsStoryboards feature launches, letting creators sketch a sequence of prompts for multi-shot videos rather than single prompts.
- Sep 30, 2025Sora 2 with audio and CameoSora 2 ships with native audio, improved physical realism, and Cameo — the ability to insert a person's likeness into generated videos with their consent.
- Oct 3, 2025Sora social appOpenAI launches a standalone Sora app with a TikTok-style feed of AI-generated short videos. Sparks immediate debate about training data and the attention-economy implications of AI slop feeds.
- Jan 20, 2026Sora Business and licensing dealsOpenAI opens Sora for commercial use with explicit rights management, announces licensing deals with several studios, and ships enterprise Sora API access.
How to access Sora
| Plan | Price | Includes |
|---|---|---|
| Sora Free | $0 | Limited Sora generations via the Sora app (invite-only in some regions). |
| ChatGPT Plus | $20/mo | Sora 2 with monthly generation credits, up to 10-second 720p outputs, standard queue. |
| ChatGPT Pro | $200/mo | Sora 2 with high generation limits, 20-second 1080p outputs, priority queue, commercial use rights. |
| Sora API | Usage-based | Sora 2 accessible via OpenAI API for developers and enterprise integration. |
Sora vs. Gemini
| Feature | Sora | Gemini |
|---|---|---|
| Model | Sora 2 | Veo 3 |
| Native audio | Yes (Sora 2) | Yes (Veo 3) |
| Max duration (consumer) | 20s (Pro tier) | 16s (Ultra tier) |
| Max resolution (consumer) | 1080p | 1080p |
| Access path | ChatGPT Plus / Pro, Sora app, API | Gemini Advanced, Vertex AI |
| Physical realism | Best in class (Sora 2) | Strong |
What Sora does well
Sora 2's core capability is producing short-form video with physical realism and multi-shot coherence that still outpaces most competitors. Ask it for "a cat knocking over a wine glass on a marble countertop, slow motion," and the physics of the glass breaking, the wine splashing, the cat's landing posture — all plausible enough to hold up in most contexts. Native audio generation (synced to the scene) came with Sora 2 and closed one of the major gaps against Veo 3.
Storyboards — the ability to sketch a sequence of prompts that Sora strings together coherently — pushed Sora past single-shot generation into something useful for short creators. And Cameo, the consented-likeness feature, opened a category of creator-led personalization that was unavailable before.
Where Sora lives
Sora ships in three places. The Sora web app at sora.com, bundled with ChatGPT Plus and Pro subscriptions. A standalone Sora mobile app with a TikTok-style feed — controversial at launch but now a meaningful consumer product. And the Sora API, launched in January 2026 for developers and enterprise integration.
The pricing structure is the same bet ChatGPT makes: most consumer users get usable generation through a $20/month subscription, power users and commercial creators subscribe to Pro at $200/month for higher limits and commercial rights, and enterprises use the API with explicit terms.
Controversies and the creator question
Sora has been the focal point of most of the industry's hardest questions about generative video. Training data sourcing has been contested — OpenAI has announced licensing deals but hasn't fully disclosed data provenance. The standalone app's AI-feed format drew criticism from creators and media observers worried about synthetic content saturation. Cameo raised consent and deepfake questions even as OpenAI built guardrails around them.
OpenAI has responded with a combination of product changes (provenance markers, consent flows, enhanced policies) and commercial moves (studio licensing, enterprise API with explicit rights management). Whether that's enough for the creator community, and whether policy responses in the US and EU change the landscape, is the story to watch in 2026.
Where Sora wins — and where it doesn't
Sora wins on raw output quality for short-form video, distribution (bundled into ChatGPT), and commercial rights clarity (especially at the Pro tier and API). For most "text in, short video out" use cases, it's the reference product.
It loses to Veo on cinematic framing for some creator workflows, to Runway on professional-grade controls and pipeline integration, and to open-weights video models on cost and flexibility for high-volume programmatic generation. For consumer use, it's the default. For specialized creative workflows, the right tool is often elsewhere in the stack.
Our coverage
Everything tagged Sora appears in Latest news above. For OpenAI broadly, see our OpenAI hub and ChatGPT hub. For the main competitor, see Gemini (which ships Veo 3). For broader industry context, see our AI news and AI analysis beats.
