When we started Hive nearly a decade ago, the world still viewed AI as promising technology with limited real world applicability. Skepticism was strong on whether or not AI models could truly be relied upon in mission-critical situations. For customers to adopt our models, the bar wasn’t simply outperforming other AI models, but detecting what humans couldn’t identify on their own.

We started with models for content moderation, drawing from our own experience in handling large scale user generated content platforms. We proved AI models could process text, image, and video content in real-time while showing human level understanding of granular classes to match moderation guidelines. Customers kept asking us to solve new challenges such as detecting IP / copyrighted content or detecting AI-generated content. Therefore, we kept expanding our proprietary model portfolio, which now currently has over 20 unique models. Presently, hundreds of the world’s largest and most innovative organizations send billions of pieces of content each day through our APIs. 

Our work did not stop there. We’ve strengthened online child safety through strategic partnerships, secured critical contracts with federal agencies, and championed responsible AI through our endorsements of the NO FAKES Act and TAKE IT DOWN Act. All the while, we have continued to invest in operating our own secure datacenter infrastructure that allows us to efficiently scale with customer demand.

We’re incredibly proud of what we’ve built over the years and the impact our technology has made in the Trust and Safety space. But just like how the AI industry is constantly evolving, so are we. 

Today, we are thrilled to announce that Hive is now Deploy. Building on our decade of expertise in developing and serving leading AI models, we’re broadening our mission as an enterprise AI inference platform, focused on democratizing high-performance frontier and best-in-class proprietary AI models.

Why Deploy?

Enterprises are shifting how they approach AI investment. The era of tokenmaxxing is over, and developers are now looking to utilize the most advanced, top-tier frontier models without breaking the bank.

Deploy is built to meet this moment. With the scale, power, and infrastructure to support high-throughput AI workloads, we unlock cost-effective optimized intelligence infrastructure for different use cases – from coding and reasoning to content safety and deepfake detection. 

“We’ve never been a company that limits itself to a single AI application. Our strength has always been our ability to develop highly specialized technology and make it operational for organizations. As the needs of enterprises and developers change, we see an opportunity to bring our skills to the open-source frontier, and we believe we are uniquely equipped to deliver.”

Kevin Guo, CEO of Deploy

Maintaining our safety ethos

While Deploy is expanding its offerings to feature open-source models, the company is still deeply committed to its safety-first ethos. Today, Hive supports many of the leading digital platforms with enterprise-grade content safety models, like Visual Moderation and CSAM Detection. As Deploy, our Content Safety suite remains a key component of our product-suite and customers can expect continued best-in-class accuracy and model performance. 

Streamlined platform experience 

As Deploy, we are not starting from scratch or moving away from what we’ve built as Hive. All existing Hive accounts and logins will remain active. 

Instead, we’ve redesigned our brand to better serve our expansion into the open-source frontier market. One of the biggest updates we’ve made is refreshing the Deploy platform, which was formerly known as Hive Portal. We’ve revamped the user experience, making it easier for customers to discover, access, and select the open source frontier and proprietary models that best meet their workload requirements. Customers can access our full portfolio of AI models and solutions in one place. 

The Deploy product suite

Customers can understand the Deploy product suite in three pillars: (1) leading LLMs, including the latest open-source models and Deploy proprietary models, (2) proprietary specialized AI models and (3) turnkey AI applications that unlock enterprise productivity. 

Leading LLMs: Developers and enterprises can access the leading open source frontier models at cost-effective prices and with the high-performance they expect. Over time, new open source frontier models will continue to become available on Deploy. Customers can also integrate Deploy VLM, which excels at multimodal classification and detection tasks, into production workflows. 

Deploy’s proprietary models are categorized into the following product suites:

• Content Safety: AI-powered content moderation solutions that analyze images, video, text, and audio to enforce content policies, automate review workflows, and support Trust & Safety operations at enterprise scale. This product suite includes Deploy’s CSAM Detection API, powered by Thorn’s trusted technology.

• AI Detection: Advanced models that identify AI-generated and deepfake content across images, video, audio, and music. Models in this suite are regularly upgraded to expand coverage to emerging generative engines. 

• IP & Publicity Detection: Specialized AI models that recognize potentially copyright-protected media and other intellectual property such as characters and brand logos, as well as detect the audio and visual likeness of public figures.

Deploy also offers turnkey applied AI software applications for enterprise productivity, such as AutoML for custom model training and fine-tuning, Hive Detect for no-code AI-generated content and deepfake detection for enterprise teams, Arbiter (formerly Moderation Dashboard) for content safety policy enforcement orchestration, and Mensio for AI-enabled sponsorship measurement & ad intelligence.

Ready to build something?

This evolution represents the future of AI at Deploy, turning ambition into execution and execution into momentum so enterprises can put AI to work quickly, safely, and cost-effectively. We hope you will join us on this new chapter, and we are excited to see what you will build. 

Explore our new website: deploy.com

Explore Models on Deploy: https://deploy.com/explore

---

About Deploy

Deploy is an intelligence infrastructure platform that gives developers and enterprises access to open source and proprietary AI models. For leading open source LLMs, Deploy operates an optimized model inference engine that keeps costs and latency low while maintaining high throughout. Deploy also offers a set of proprietary AI models that address challenges in content safety, AI & deepfake detection, sponsorship intelligence, and IP & publicity protection. Trusted by the world’s largest and most innovative organizations, Deploy’s solutions unlock high-performance AI model inference at production scale. 

Deploy has raised more than $200M from leading investors, including Tomales Bay Capital, General Catalyst, 8VC, Glynn Capital, Bain & Company, Visa Ventures, and others. The company has more than 250 employees across offices in San Francisco, Seattle, New York, and Delhi.