Microsoft has launched Bing Video Creator, integrating OpenAI’s advanced Sora model to provide free text-to-video generation. This debut significantly broadens public access to sophisticated AI video tools, previously often locked behind substantial paywalls. The company detailed that the service aims to empower users by transforming text prompts into short videos. This strategic move positions Bing to attract users by offering a cutting-edge AI feature without direct cost, challenging existing paid services.
Initially, Bing Video Creator is accessible globally (excluding China and Russia) through the Bing mobile app, with future rollouts planned for the Bing desktop experience and Copilot Search. Users receive ten “Fast” video creations before needing to redeem 100 Microsoft Rewards points for further expedited generations or continuing at standard speeds.
This free access model contrasts with OpenAI’s own Sora offering, which typically requires a ChatGPT Plus subscription starting around $20 per month. According to the Bing Blogs, the platform’s goal is to make creativity “effortless and accessible.”
Currently, generated videos are five seconds long and presented in a 9:16 aspect ratio, with a 16:9 format expected later. Users can queue up to three video generations simultaneously, and their creations are stored for up to 90 days. Desktop access will eventually be available via Bing.com/create. During a beta phase in late May, Bing Video Creator generated videos at 480p, though the official launch did not specify output resolution.
Democratizing Advanced AI Video Generation
Microsoft’s introduction of the free Bing Video Creator signals a notable shift in the AI video generation landscape. Previously, access to powerful models like OpenAI’s Sora, launched for ChatGPT Plus and Pro subscribers in December 2024, was primarily through paid tiers.
Similarly, Google’s Veo 3 model, unveiled in May with features like integrated audio, voice lip-sinc, is available via its new premium AI Ultra plan.
This launch continues Microsoft’s pattern of embedding OpenAI’s technology across its product suite. Bing Image Creator, utilizing DALL-E, was introduced two years ago, with DALL-E 3 later integrated into Microsoft Designer in November 2023 and Bing Image Creator around September 2023. These earlier efforts paved the way for offering Sora’s video capabilities.
Navigating Quality, Features, and Responsible AI
While democratizing access is a key theme, potential users should be aware of the tool’s current capabilities and Microsoft’s past experiences. Reports from the beta period indicated that Bing Video Creator then only supported text-to-video and did not allow image uploads to generate videos, a feature present in Sora via ChatGPT.
Microsoft acknowledged during the beta, as per a statement cited by Windows Latest, that “While we’re actively improving the experience, you may occasionally notice limited functionality—such as your video history not saving as expected.” The standard, free generation mode was also reported to take a “few hours” during beta.
Microsoft has emphasized its commitment to responsible AI. The Bing Video Creator incorporates existing Sora safeguards alongside additional protections to limit harmful content. Furthermore, like Bing Image Creator, it will use C2PA standard content credentials to help identify AI-generated videos.
This is crucial, especially given past challenges with AI integrations. In January 2025 that Microsoft had to revert a Bing Image Creator DALL-E update due to quality issues, and earlier, in October 2023, faced issues with erroneous content violation flags.
The decision to launch Bing Video Creator initially on mobile, and not yet within the dedicated Microsoft Copilot tool, has drawn some commentary. Windows Central suggested this could be a strategy to boost Bing app usage. The AI video generation field is intensely competitive.
Beyond OpenAI and Google, companies like Adobe with its Firefly Video Model and Runway ML with Gen-3 Alpha Turbo offer robust solutions. Moreover, the open-source movement, exemplified by Alibaba’s free Wan 2.1 AI video models released in February, presents another dynamic.
On the Artificial Analysis Video Arena Leaderboard Google’s Veo 3 is currently ranked on top before OpenAI’s Sora, indicating the rapid evolution and varying performance benchmarks in this space.
