Akamai launches AI inference cloud with 12 open-weight models

Akamai has launched Akamai Inference, a cloud AI inference service offering 12 open-weight models ranging from eight billion to 2.8 trillion parameters. The Nasdaq-listed company announced the beta-stage offering this week alongside Akamai Agent Sandbox and an expanded dedicated GPU fleet.

The new service gives enterprises managed access to open-weight and custom models through OpenAI-compatible APIs (interfaces that let developers swap model providers without rewriting code). Customers choose between two consumption models:

• Serverless, pay-per-use inference — no infrastructure management
• Dedicated deployments — reserved capacity on Akamai hardware

The launch lineup includes Nvidia Nemotron 3 Ultra, OpenAI gpt-oss-120b, and Google Gemma 4, plus embedding models for retrieval-augmented generation (RAG, a technique that grounds model answers in a company's own documents).

What does the sandbox do?

Akamai Agent Sandbox addresses a specific agentic problem: AI agents that write and execute their own code need somewhere contained to run it. The new service gives agents and the code they generate a restricted execution environment, keeping that workload isolated from production systems.

COO and GM of Akamai Cloud Adam Karon framed the launch as a response to customer demand. "We have been implementing an AI infrastructure strategy based on the notion that AI inference will represent the majority of enterprise focus and investment. Our customers proved that thesis and pulled us further," he said.

"They want to build, run, and protect AI on the same platform," Karon continued. "Akamai Inference and Akamai Agent Sandbox add the inference and execution layers that turn our AI infrastructure into a full AI stack, and our security portfolio protects every layer of it."

Hardware under the hood

The GPU expansion has two tiers. Akamai's fleet now includes Nvidia RTX Pro 6000 Blackwell Server Edition GPUs positioned for low-latency inference, and Blackwell Ultra GPU clusters aimed at enterprises that want to self-host models on infrastructure they control. The company is also preparing its data centers to support the Nvidia Vera Rubin platform next year.

Distribution is the other differentiator Akamai claims. Akamai Cloud can now offer GPU and CPU clusters in core regions plus more than 4,350 edge locations, putting inference compute physically closer to end users.

What does this change for buyers?

SVP of cloud technology Jon Alexander said the pitch is about removing an either-or choice. "Enterprises should not have to choose between speed and control when they put AI into production," he said.

"Some teams want to ship an AI feature this quarter on a managed platform. Others need dedicated GPUs to run frontier-class open models on their own terms. Every team running agents needs a boundary around the code those agents write," Alexander said. "Akamai Cloud now supports all of them on one platform, with security and resilience built in."

Akamai Inference is available in beta now. Pricing follows the serverless pay-per-use model or dedicated deployment terms depending on the chosen tier.