Amazon CEO Andy Jassy outlined his vision for the artificial intelligence market during the company's Q3 2026 earnings call Thursday, forecasting that the competition among frontier models will produce a handful of viable options rather than consolidating around a single leader. Jassy told analysts that the real challenge facing enterprises will shift from choosing a single model to selecting among several capable alternatives.
When responding to a Goldman Sachs analyst's question about Amazon Bedrock's market potential, Jassy projected that within two to three years, "6- 8 models that are about as good as each other, or at least good enough for 95%+ of the workloads" will dominate the landscape.
No single AI winner
Jassy drew parallels to earlier technology cycles where early leaders eventually faced serious competition. He cited databases and analytics platforms as historical precedents, along with TensorFlow, which appeared destined to become the standard machine learning framework before PyTorch emerged as a stronger contender.
The CEO expects artificial intelligence models to follow a similar pattern. Different models will excel in distinct areas—some stronger at reasoning or code generation, others offering advantages in visual processing, latency, or cost efficiency. This specialization creates strong incentives for organizations to deploy multiple models rather than standardize on a single provider. According to Jassy, "most customers are going to use multiple models, and many will use multiple models in the same generative AI application." This approach allows development teams to select the optimal model for each specific task rather than locking into one vendor's entire portfolio.
One API, many models
Bedrock's architecture directly reflects this multi-model philosophy. The service provides access to more than 100 foundation models from vendors including Anthropic, OpenAI, Meta, Mistral, DeepSeek, Moonshot AI, and xAI through a single API interface. This design lets teams invoke Claude, GPT, Llama, or specialized smaller models without building separate integration pipelines for each.
Amazon recently expanded Bedrock to include OpenAI's models alongside the proprietary models it previously hosted. The company has also integrated managed versions of open-source models such as Moonshot's Kimi and Zhipu's GLM, reflecting what Jassy characterizes as sustained demand for Western open-source options that Bedrock will continue to satisfy.
Switching models isn't free
While Bedrock's Converse API provides a standardized interface for working with models from different providers without rewriting integrations, model switching carries practical constraints. Each model operates with different context windows, tool support, and feature sets, requiring developers to validate application behavior across different options.
Complexity increases when teams have structured their workflows around a particular vendor's agent framework. Jassy acknowledged that customers want to "swap them back and forth as they leapfrog each other, as different constraints of their applications require them."
Data protection considerations also influence these decisions. AWS maintains that Bedrock prevents model providers from accessing customer prompts or responses, and that this data does not feed into model training. Jassy highlighted these safeguards as a significant factor driving enterprise adoption of the service.
Why Amazon wins anyway
Jassy's strategy rests on a fundamental insight: Amazon does not need to build the best model if it controls the platform where developers choose among them. This positioning is not unique—Google's Vertex AI and Microsoft's Azure AI Foundry pursue similar multi-provider strategies with unified access layers—but it represents a significant business opportunity.
A market where six to eight models continuously compete and teams regularly rotate between them creates ideal conditions for a neutral access layer to capture value. Jassy makes little attempt to conceal that Bedrock is the intended beneficiary, describing the opportunity as "gigantic" and suggesting the service could eventually rival EC2, which remains one of AWS's largest revenue generators.
Early performance metrics support this trajectory. Fortune 500 companies are now spending nearly seven times more on Bedrock compared to the prior year, token volume has grown almost 17 times, and Bedrock ranks as the fastest-growing multibillion-dollar product in AWS history by Jassy's account.
This momentum arrives as AWS accelerates overall, with the segment expanding 39.3% in the quarter to $46 billion and reaching a $184 billion annualized run rate. Jassy noted that the majority of Bedrock traffic already runs on Amazon's proprietary Trainium chips, a position that allows the company to maintain competitive pricing while protecting profitability.