Amazon is working on Olympus, a new multimodal AI system centered on image and video analysis. The project points to a clear strategic choice: instead of competing only on text processing, Amazon is looking for ground where visual AI can create practical value for AWS customers.
Why Olympus matters for Amazon AI
According to The Information, citing multiple sources familiar with the project, Olympus is designed to focus on visual understanding. That emphasis matters because Amazon has limitations in text processing and solving complex problems compared with models from OpenAI and Anthropic.
The reported goal is not just to add a basic video feature to an existing AI product. Sources say Olympus goes much further, with capabilities aimed at detailed analysis of images and video. In a market where major cloud providers need reasons for customers to choose their platforms, that kind of specialization could become important.
Amazon is already tied closely to Anthropic after putting $8 billion into the company. Even so, the source article makes clear why Amazon would keep building its own systems: exclusive AI offerings can help cloud providers stand apart while still offering outside models.
Video search is the core opportunity
One of the most direct use cases for Olympus is high-accuracy video archive search. That capability could matter for media companies managing large video libraries, where finding the right moment inside footage can be difficult if the system only understands file names, rough metadata or manual tags.
Olympus is also described as useful for sports footage. Sources say the system can track movement with high accuracy, including the exact moment a basketball leaves a player's hand, and estimate the trajectory. That kind of analysis could replace work typically done by human analysts.
The examples in the source point to a broader pattern. Olympus is being framed as a system for understanding what happens inside video, not merely labeling a clip at a high level. For customers, the value would come from locating, following and interpreting events inside visual material.
Industrial uses could extend beyond media
The reported uses are not limited to sports analytics firms and media companies. Industrial applications could include AI-powered inspections of underwater drilling equipment to find corrosion or leaks.
That example shows why video-focused AI can be more than a content tool. In industrial settings, visual inspection often depends on recognizing specific changes or signs of damage. A system that can search footage and analyze visual details could be relevant wherever organizations collect video for review.
The source does not provide technical details on how Olympus works. For now, the important point is the direction: Amazon appears to be building a visual model strategy that can serve concrete workflows in archives, sports analysis and inspections.
How Olympus fits with Amazon's model plans
Olympus is not the only AI work described in the report. Late last year, Amazon's AI chief Rohit Prasad laid out plans for four large text models, sources told The Information.
Two of those models stand out in the source: one with 400 billion parameters and another with two trillion parameters, larger than the original GPT-4. The article also notes that parameter count matters less now because of new efficiency techniques, though it still gives a rough measure of model capability.
At that time, Amazon planned to pair the text models with a smaller visual processing model. The exact relationship between those plans and Olympus remains unclear because technical details are still under wraps.
Amazon may reveal more at next week's AWS re:Invent cloud conference. Until then, the public picture is limited to what sources have described: a multimodal AI system with a visual focus, connected to Amazon's broader effort to improve its AI position.
The AWS strategy behind the project
The cloud market context is central to the Olympus story. Microsoft works with OpenAI, Google builds Gemini, and Amazon wants its own edge in the market. For AWS, a distinctive AI system could help make the platform more attractive to customers deciding where to build or run AI workloads.
The source also connects Amazon's model work to its chip strategy. Like Google, Microsoft and Meta, Amazon is developing its own AI chips to reduce dependence on Nvidia's pricing control and chip shortages in the AI chip market.
With its own models and models from Anthropic, Amazon would have clear use cases for those custom chips. Olympus could therefore serve more than one purpose: it could expand Amazon's AI product lineup, strengthen AWS differentiation and give Amazon additional workloads for its own chip investments.
The larger takeaway is straightforward. Amazon is not relying on a single AI bet. It is backing Anthropic, planning large text models, developing custom chips and building Olympus around video and image analysis. If the system performs as sources describe, Olympus could become one of Amazon's more distinctive answers to rivals in the AI cloud race.