{"success":true,"data":{"pressRelease":{"id":"180690","rtpr_id":"nGNX59MJ47-20261027","ticker":"AKAM","exchange":"NASDAQ","all_tickers":["AKAM"],"title":"Akamai Launches AI Inference Service and Agent Execution Environment","author":"Globe Newswire","published_at":"2026-10-27T10:30:00.262Z","article_body":"CAMBRIDGE, Mass., Oct. 27, 2026 (GLOBE NEWSWIRE) -- Bringing managed model\nserving and isolated execution for AI agents to Akamai Cloud, Akamai (Nasdaq:\nAKAM) today introduced two foundational products for its AI stack: Akamai\nInference and Akamai Agent Sandbox. With these additions, enterprises can\nbuild and tune their own or hosted open-weight models, ground them in their\nown data, run agents inside enforced boundaries, and secure the applications,\nagents, and infrastructure involved, all on one cloud platform with\nAI-optimized infrastructure.\n\nThese capabilities mark the latest step in Akamai’s AI strategy: giving\nenterprises the trusted global infrastructure and services required to serve,\nrun, and secure AI applications and agents, close to the users and data they\ndepend on. Akamai Inference and Akamai Agent Sandbox now add managed services\non top of that infrastructure — a foundation that has already secured\nmultiyear contracts for Akamai Cloud collectively worth more than $14 billion.\nThe momentum highlights how Akamai’s ability to leverage both CPUs and GPUs\n— together with cloud storage, networking, and edge functions on its\ninterconnected global network fabric — is quickly becoming the trusted\nfoundation for customers to build, deploy, and operate AI workloads at scale.\n\nEnterprises moving AI into production need far more than models. They need\nserving infrastructure, access to their data, routing and governance, and, as\nagents begin to write and execute their own code, a hard boundary around what\nthat code can reach. Most teams assemble those pieces from separate providers\nand secure the interaction points themselves. Akamai Cloud now delivers a\nconnected AI environment, placed across a continuum of compute that runs from\ndense GPU and CPU clusters in core regions to more than 4,350 edge locations,\nand protected by the same security portfolio that defends many of the\nworld’s largest online businesses.\n\n“We have been implementing an AI infrastructure strategy based on the notion\nthat AI inference will represent the majority of enterprise focus and\ninvestment. Our customers proved that thesis and pulled us further,” said\nAdam Karon, Chief Operating Officer and GM of Akamai Cloud. “They want to\nbuild, run, and protect AI on the same platform. Akamai Inference and Akamai\nAgent Sandbox add the inference and execution layers that turn our AI\ninfrastructure into a full AI stack, and our security portfolio protects every\nlayer of it. Akamai is uniquely positioned to bring our scale and distribution\nto address scaling real-time inference and agentic workloads.”\n\nAccess models without operating the stack: Akamai Inference\n\nAkamai Inference gives enterprises managed access to open-weight and custom\nmodels through OpenAI-compatible APIs, so existing applications move without\nrewrites. Customers choose serverless, pay-per-use inference or dedicated\ndeployments. At launch, the catalog includes 12 open-weight models ranging\nfrom 8 billion to 2.8 trillion parameters, including NVIDIA(™) Nemotron 3\nUltra, OpenAI(™) gpt-oss-120b, and Google(™) Gemma 4, along with embedding\nmodels for retrieval-augmented generation.\n\nAkamai Cloud provides access to managed databases and object storage, and\nseamlessly connects to private and public cloud platforms, so models can draw\non enterprise data. Akamai Inference then handles the orchestration, model\nexecution, and authentication. Requests run on the Akamai global network,\nwhich reduces latency, and helps to ensure high availability and consistent\nperformance while maintaining predictable service levels. \n\nGive agents a place to execute: Akamai Agent Sandbox\n\nAI agents are increasingly doing real work, from taking actions on a user’s\nbehalf to writing and running their own code. Akamai Agent Sandbox gives\ndevelopers a secure place to run both. Agents run on Akamai Functions, and the\ncode they write runs in its own isolated environment, sequestered from the\nagent’s runtime. Delivered as an open source component of the Spin\nframework, Akamai Agent Sandbox uses WebAssembly isolation to set clear limits\non compute, memory outbound connections, and system capabilities. Developers\ncan deploy agents with confidence about where they run and what they can\nreach.\n\nAgent isolation is converging on several models, from microVMs with snapshot\nand fork capability, to policy-enforced containers, to WebAssembly-isolated\nfunctions. Akamai allows developers to choose the architecture that fits their\nworkload, with NVIDIA-accelerated and dedicated CPU infrastructure, managed\nKubernetes, and Akamai Functions for serverless execution across edge and\ncore. Akamai also plans to add capabilities that govern what agents can reach\nand what can reach them.\n\nBuild and tune on a broader GPU fleet: NVIDIA Blackwell Ultra today, Vera\nRubin next\n\nAkamai also expanded its dedicated GPU fleet to include NVIDIA Blackwell Ultra\nGPU clusters and is preparing its infrastructure to support the NVIDIA Vera\nRubin platform next year. Akamai’s GPU fleet now spans NVIDIA RTX PRO\n(https://www.globenewswire.com/Tracker?data=GG0S-m-scJqvuVUF9sD_MNDkRRpxH8HOryehHUn_1E_SPAbFJ2Z2bwKiU7PyHlf4Av9FOKf5m4znQXkC26MpmMqCePd-TZJLP8U4G5eYqJ8dTKcFvJdmFEuOppBJFwTd1HlOnWuTTIsabK-IJn3JcBONpy_jU0tSWtiXJNTkYlo=)(™)\n6000 Blackwell Server Edition\n(https://www.globenewswire.com/Tracker?data=auvJDoVzuPkdKHtVrCrWlQMbOPX6iuX3osgynWhWMUtKaDLPImhveJKBBJVDqAzWGlw9hHfr8JlEFem8WRqEst1164P3K2rHyWCqijQWY_lXi2Kc7VbSCtQCJHXFTDUkpBeupfa2ZK5eSL02WPvYa-cE5kXlhEdrIr_-AbwSvVlTBxXwRk57hWhrm7AfNlhQ)(®)\nGPUs for low-latency inference and NVIDIA Blackwell Ultra GPU clusters for\nenterprises that want to build, tune, and self-host models on infrastructure\nthey control. Akamai is also preparing its data centers to support the NVIDIA\nVera Rubin platform in 2027.\n\nEach eight-GPU Blackwell Ultra node carries 288 GB of HBM3e memory per GPU,\n2.3 TB in total, enough to hold frontier-class models. The Blackwell Ultra\narchitecture targets the workloads driving that memory demand: test-time\nscaling, chain-of-thought reasoning, and long-context agentic work. GPU\nclusters give enterprises control over model versions and deprecation\ntimelines, and run without rate limits or shared-tenancy queueing. They\nconnect to the same data, networking, and security services as the rest of\nAkamai Cloud.\n\n“Enterprises should not have to choose between speed and control when they\nput AI into production,” said Jon Alexander, Senior Vice President of Cloud\nTechnology at Akamai. “Some teams want to ship an AI feature this quarter on\na managed platform. Others need dedicated GPUs to run frontier-class open\nmodels on their own terms. Every team running agents needs a boundary around\nthe code those agents write. Akamai Cloud now supports all of them on one\nplatform, with security and resilience built in.”\n\nAvailability\nAkamai Agent Sandbox (via Spin SDK) and Akamai Inference are available in\nbeta. Akamai Inference beta includes serverless deployment options. Contact\nAkamai Sales for NVIDIA B300 GPU clusters availability.\n\nAbout Akamai\n\nAkamai is the cloud company that powers and protects an AI-driven world. Our\ncloud platform extends high-performance cloud computing from the core to the\nedge, enabling organizations to build and scale next-generation AI\napplications and providing comprehensive, multilayered security to safeguard\nenterprises against evolving cyberthreats. Learn more at akamai.com\n(https://www.globenewswire.com/Tracker?data=8IFcDBi8ieDJYkLov6lOI8GcqL_4f6mbqgo9SqYS14qp6IRe0qxXNUnIDINHGtRwqyUa3JqJ_zg8dTJYqvA8LQ==)\nand akamai.com/blog\n(https://www.globenewswire.com/Tracker?data=8IFcDBi8ieDJYkLov6lOI365PD6AYBXY1C-oniQ5ri-9k9zgA7mlryWqGxWuu69FdcFzVJSnrTG9z0anMV9n6IrlGk7pAyXFmzY9WenVT_8=),\nor follow Akamai Technologies on X\n(https://www.globenewswire.com/Tracker?data=DCpgllsrw1Xu385LnvRjqY7JXN_zB6CurJQYA1swXlkV5F0T9TQNTksM8zz9Lv79G-AEp5uscQU4Lz_0lU9rfg==)\nand LinkedIn\n(https://www.globenewswire.com/Tracker?data=NAKeN0LJlaGtc9z2Zg8hlOOgBr3ayRbaT2JevWNIiPM6KxSLfO2kQLP0VsEMy4ZnV4kSPOZvkrPfCxDg05nJaeG2F9FAN5H-ACyHuKLK4z-4jEU-IV1tgVIvryUFXB7I).\n\nAkamai Statement Under the Private Securities Litigation Reform Act\n\nThis press release contains statements that are not statements of historical\nfact and constitute forward-looking statements for purposes of the safe harbor\nprovisions under The Private Securities Litigation Reform Act of 1995,\nincluding, but not limited to, statements about Akamai Inference Cloud, its\nanticipated capabilities, scalability, performance, global deployment plans\nand the expected benefits to Akamai, current and prospective customers and end\nusers. Each of the forward-looking statements is subject to change as a result\nof various important factors, many of which are beyond Akamai’s control,\nincluding, but not limited to: Akamai's inability to achieve the expected\nperformance or benefits of Akamai Inference Cloud; Akamai’s capabilities\nfailing to meet expectations, including due to defects, security breaches,\ndelays in performance, challenges in leveraging NVIDIA's technologies, or\nother similar problems; effects of competition, including pricing pressure and\nchanging business models; changes in customer or user preferences or demands;\nmacroeconomic trends and uncertainties, including industry volatility, the\neffects of inflation, fluctuating interest and foreign currency exchange\nrates, supply chain and logistics costs, constraints, changes or disruptions;\ndefects or disruptions in Akamai’s or NVIDIA's products or IT systems,\nincluding cyber-attacks, data breaches or malware; changes to economic,\npolitical and regulatory conditions in the United States or internationally;\nand other factors that are discussed in Akamai’s most recent annual report\non Form 10-K, subsequent quarterly reports on Form 10-Q and other documents\nfiled with the Securities and Exchange Commission. The forward-looking\nstatements contained herein are not guarantees of future performance and speak\nonly as of the date hereof, and, except as required by law, Akamai disclaims\nany obligation to update these forward-looking statements to reflect future\nevents or circumstances.\n\nContacts\nAkamai Media Relations\nakamaipr@akamai.com\n\nAkamai Investor Relations\ninvrel@akamai.com\n\n(https://www.globenewswire.com/NewsRoom/AttachmentNg/4ce7685a-3da0-416d-9bc1-58cc8b7812b8)\n\n\n\nGlobeNewswire, Inc. 2026","article_body_html":"","raw_payload":{"data":{"id":"nGNX59MJ47-20261027","title":"Akamai Launches AI Inference Service and Agent Execution Environment","author":"Globe Newswire","ticker":"AKAM","created":"2026-10-27T10:30:00.262Z","tickers":["AKAM"],"exchange":"NASDAQ","article_body":"CAMBRIDGE, Mass., Oct. 27, 2026 (GLOBE NEWSWIRE) -- Bringing managed model\nserving and isolated execution for AI agents to Akamai Cloud, Akamai (Nasdaq:\nAKAM) today introduced two foundational products for its AI stack: Akamai\nInference and Akamai Agent Sandbox. With these additions, enterprises can\nbuild and tune their own or hosted open-weight models, ground them in their\nown data, run agents inside enforced boundaries, and secure the applications,\nagents, and infrastructure involved, all on one cloud platform with\nAI-optimized infrastructure.\n\nThese capabilities mark the latest step in Akamai’s AI strategy: giving\nenterprises the trusted global infrastructure and services required to serve,\nrun, and secure AI applications and agents, close to the users and data they\ndepend on. Akamai Inference and Akamai Agent Sandbox now add managed services\non top of that infrastructure — a foundation that has already secured\nmultiyear contracts for Akamai Cloud collectively worth more than $14 billion.\nThe momentum highlights how Akamai’s ability to leverage both CPUs and GPUs\n— together with cloud storage, networking, and edge functions on its\ninterconnected global network fabric — is quickly becoming the trusted\nfoundation for customers to build, deploy, and operate AI workloads at scale.\n\nEnterprises moving AI into production need far more than models. They need\nserving infrastructure, access to their data, routing and governance, and, as\nagents begin to write and execute their own code, a hard boundary around what\nthat code can reach. Most teams assemble those pieces from separate providers\nand secure the interaction points themselves. Akamai Cloud now delivers a\nconnected AI environment, placed across a continuum of compute that runs from\ndense GPU and CPU clusters in core regions to more than 4,350 edge locations,\nand protected by the same security portfolio that defends many of the\nworld’s largest online businesses.\n\n“We have been implementing an AI infrastructure strategy based on the notion\nthat AI inference will represent the majority of enterprise focus and\ninvestment. Our customers proved that thesis and pulled us further,” said\nAdam Karon, Chief Operating Officer and GM of Akamai Cloud. “They want to\nbuild, run, and protect AI on the same platform. Akamai Inference and Akamai\nAgent Sandbox add the inference and execution layers that turn our AI\ninfrastructure into a full AI stack, and our security portfolio protects every\nlayer of it. Akamai is uniquely positioned to bring our scale and distribution\nto address scaling real-time inference and agentic workloads.”\n\nAccess models without operating the stack: Akamai Inference\n\nAkamai Inference gives enterprises managed access to open-weight and custom\nmodels through OpenAI-compatible APIs, so existing applications move without\nrewrites. Customers choose serverless, pay-per-use inference or dedicated\ndeployments. At launch, the catalog includes 12 open-weight models ranging\nfrom 8 billion to 2.8 trillion parameters, including NVIDIA(™) Nemotron 3\nUltra, OpenAI(™) gpt-oss-120b, and Google(™) Gemma 4, along with embedding\nmodels for retrieval-augmented generation.\n\nAkamai Cloud provides access to managed databases and object storage, and\nseamlessly connects to private and public cloud platforms, so models can draw\non enterprise data. Akamai Inference then handles the orchestration, model\nexecution, and authentication. Requests run on the Akamai global network,\nwhich reduces latency, and helps to ensure high availability and consistent\nperformance while maintaining predictable service levels. \n\nGive agents a place to execute: Akamai Agent Sandbox\n\nAI agents are increasingly doing real work, from taking actions on a user’s\nbehalf to writing and running their own code. Akamai Agent Sandbox gives\ndevelopers a secure place to run both. Agents run on Akamai Functions, and the\ncode they write runs in its own isolated environment, sequestered from the\nagent’s runtime. Delivered as an open source component of the Spin\nframework, Akamai Agent Sandbox uses WebAssembly isolation to set clear limits\non compute, memory outbound connections, and system capabilities. Developers\ncan deploy agents with confidence about where they run and what they can\nreach.\n\nAgent isolation is converging on several models, from microVMs with snapshot\nand fork capability, to policy-enforced containers, to WebAssembly-isolated\nfunctions. Akamai allows developers to choose the architecture that fits their\nworkload, with NVIDIA-accelerated and dedicated CPU infrastructure, managed\nKubernetes, and Akamai Functions for serverless execution across edge and\ncore. Akamai also plans to add capabilities that govern what agents can reach\nand what can reach them.\n\nBuild and tune on a broader GPU fleet: NVIDIA Blackwell Ultra today, Vera\nRubin next\n\nAkamai also expanded its dedicated GPU fleet to include NVIDIA Blackwell Ultra\nGPU clusters and is preparing its infrastructure to support the NVIDIA Vera\nRubin platform next year. Akamai’s GPU fleet now spans NVIDIA RTX PRO\n(https://www.globenewswire.com/Tracker?data=GG0S-m-scJqvuVUF9sD_MNDkRRpxH8HOryehHUn_1E_SPAbFJ2Z2bwKiU7PyHlf4Av9FOKf5m4znQXkC26MpmMqCePd-TZJLP8U4G5eYqJ8dTKcFvJdmFEuOppBJFwTd1HlOnWuTTIsabK-IJn3JcBONpy_jU0tSWtiXJNTkYlo=)(™)\n6000 Blackwell Server Edition\n(https://www.globenewswire.com/Tracker?data=auvJDoVzuPkdKHtVrCrWlQMbOPX6iuX3osgynWhWMUtKaDLPImhveJKBBJVDqAzWGlw9hHfr8JlEFem8WRqEst1164P3K2rHyWCqijQWY_lXi2Kc7VbSCtQCJHXFTDUkpBeupfa2ZK5eSL02WPvYa-cE5kXlhEdrIr_-AbwSvVlTBxXwRk57hWhrm7AfNlhQ)(®)\nGPUs for low-latency inference and NVIDIA Blackwell Ultra GPU clusters for\nenterprises that want to build, tune, and self-host models on infrastructure\nthey control. Akamai is also preparing its data centers to support the NVIDIA\nVera Rubin platform in 2027.\n\nEach eight-GPU Blackwell Ultra node carries 288 GB of HBM3e memory per GPU,\n2.3 TB in total, enough to hold frontier-class models. The Blackwell Ultra\narchitecture targets the workloads driving that memory demand: test-time\nscaling, chain-of-thought reasoning, and long-context agentic work. GPU\nclusters give enterprises control over model versions and deprecation\ntimelines, and run without rate limits or shared-tenancy queueing. They\nconnect to the same data, networking, and security services as the rest of\nAkamai Cloud.\n\n“Enterprises should not have to choose between speed and control when they\nput AI into production,” said Jon Alexander, Senior Vice President of Cloud\nTechnology at Akamai. “Some teams want to ship an AI feature this quarter on\na managed platform. Others need dedicated GPUs to run frontier-class open\nmodels on their own terms. Every team running agents needs a boundary around\nthe code those agents write. Akamai Cloud now supports all of them on one\nplatform, with security and resilience built in.”\n\nAvailability\nAkamai Agent Sandbox (via Spin SDK) and Akamai Inference are available in\nbeta. Akamai Inference beta includes serverless deployment options. Contact\nAkamai Sales for NVIDIA B300 GPU clusters availability.\n\nAbout Akamai\n\nAkamai is the cloud company that powers and protects an AI-driven world. Our\ncloud platform extends high-performance cloud computing from the core to the\nedge, enabling organizations to build and scale next-generation AI\napplications and providing comprehensive, multilayered security to safeguard\nenterprises against evolving cyberthreats. Learn more at akamai.com\n(https://www.globenewswire.com/Tracker?data=8IFcDBi8ieDJYkLov6lOI8GcqL_4f6mbqgo9SqYS14qp6IRe0qxXNUnIDINHGtRwqyUa3JqJ_zg8dTJYqvA8LQ==)\nand akamai.com/blog\n(https://www.globenewswire.com/Tracker?data=8IFcDBi8ieDJYkLov6lOI365PD6AYBXY1C-oniQ5ri-9k9zgA7mlryWqGxWuu69FdcFzVJSnrTG9z0anMV9n6IrlGk7pAyXFmzY9WenVT_8=),\nor follow Akamai Technologies on X\n(https://www.globenewswire.com/Tracker?data=DCpgllsrw1Xu385LnvRjqY7JXN_zB6CurJQYA1swXlkV5F0T9TQNTksM8zz9Lv79G-AEp5uscQU4Lz_0lU9rfg==)\nand LinkedIn\n(https://www.globenewswire.com/Tracker?data=NAKeN0LJlaGtc9z2Zg8hlOOgBr3ayRbaT2JevWNIiPM6KxSLfO2kQLP0VsEMy4ZnV4kSPOZvkrPfCxDg05nJaeG2F9FAN5H-ACyHuKLK4z-4jEU-IV1tgVIvryUFXB7I).\n\nAkamai Statement Under the Private Securities Litigation Reform Act\n\nThis press release contains statements that are not statements of historical\nfact and constitute forward-looking statements for purposes of the safe harbor\nprovisions under The Private Securities Litigation Reform Act of 1995,\nincluding, but not limited to, statements about Akamai Inference Cloud, its\nanticipated capabilities, scalability, performance, global deployment plans\nand the expected benefits to Akamai, current and prospective customers and end\nusers. Each of the forward-looking statements is subject to change as a result\nof various important factors, many of which are beyond Akamai’s control,\nincluding, but not limited to: Akamai's inability to achieve the expected\nperformance or benefits of Akamai Inference Cloud; Akamai’s capabilities\nfailing to meet expectations, including due to defects, security breaches,\ndelays in performance, challenges in leveraging NVIDIA's technologies, or\nother similar problems; effects of competition, including pricing pressure and\nchanging business models; changes in customer or user preferences or demands;\nmacroeconomic trends and uncertainties, including industry volatility, the\neffects of inflation, fluctuating interest and foreign currency exchange\nrates, supply chain and logistics costs, constraints, changes or disruptions;\ndefects or disruptions in Akamai’s or NVIDIA's products or IT systems,\nincluding cyber-attacks, data breaches or malware; changes to economic,\npolitical and regulatory conditions in the United States or internationally;\nand other factors that are discussed in Akamai’s most recent annual report\non Form 10-K, subsequent quarterly reports on Form 10-Q and other documents\nfiled with the Securities and Exchange Commission. The forward-looking\nstatements contained herein are not guarantees of future performance and speak\nonly as of the date hereof, and, except as required by law, Akamai disclaims\nany obligation to update these forward-looking statements to reflect future\nevents or circumstances.\n\nContacts\nAkamai Media Relations\nakamaipr@akamai.com\n\nAkamai Investor Relations\ninvrel@akamai.com\n\n(https://www.globenewswire.com/NewsRoom/AttachmentNg/4ce7685a-3da0-416d-9bc1-58cc8b7812b8)\n\n\n\nGlobeNewswire, Inc. 2026"},"type":"article","timestamp":"2026-10-27T10:30:00.318785033Z","server_sent_at_ms":1793097000318},"received_at":"2026-10-27T10:30:00.407Z","source_url":null},"analysis":{"id":"168525","press_release_id":"180690","analysis_json":{"industry":{"label":"IT Services","sector":"Information Technology"},"redFlags":["products are beta-stage; revenue impact not yet quantified","forward-looking statements flag dependence on NVIDIA technology and competitive pricing pressure"],"eventType":"product_launch","narrative":"Akamai launched two new AI stack products, Akamai Inference and Akamai Agent Sandbox, bringing managed model serving and isolated agent execution to Akamai Cloud.\n\nThe launch builds on Akamai Cloud infrastructure backed by multiyear contracts worth more than $14 billion, spanning over 4,350 edge locations and an expanded GPU fleet that now includes NVIDIA Blackwell Ultra clusters, with Vera Rubin support planned for 2027.\n\nBoth products are available in beta; the move positions Akamai to capture enterprise AI inference and agentic workload revenue on a single platform, though revenue contribution remains early-stage.","sentiment":"bullish","agentHooks":{"shouldPost":true,"suggestedAngle":"Akamai extends its edge/cloud franchise into managed AI inference and agent sandboxing -- a credible challenge to hyperscaler AI stacks built on a $14B+ cloud contract base."},"keyFigures":{"customDimensions":{"edge_locations":"more than 4,350","model_param_range":"8 billion to 2.8 trillion parameters","models_in_catalog":12,"gpu_memory_per_node":"2.3 TB total (288 GB HBM3e per GPU)","cloud_contracts_value":"more than $14 billion"}},"quotedText":"They want to build, run, and protect AI on the same platform. Akamai Inference and Akamai Agent Sandbox add the inference and execution layers that turn our AI infrastructure into a full AI stack","namedEntities":{"people":[{"name":"Adam Karon","role":"Chief Operating Officer and GM of Akamai Cloud"},{"name":"Jon Alexander","role":"Senior Vice President of Cloud Technology"}],"products":["Akamai Inference","Akamai Agent Sandbox","Akamai Cloud","Akamai Functions","NVIDIA Nemotron 3 Ultra","OpenAI gpt-oss-120b","Google Gemma 4","NVIDIA Blackwell Ultra","NVIDIA RTX PRO 6000 Blackwell Server Edition","Spin framework"],"companies":[{"name":"Akamai","ticker":"AKAM","relationship":"filer"},{"name":"NVIDIA","relationship":"supplier"},{"name":"OpenAI","relationship":"model provider"},{"name":"Google","relationship":"model provider"}],"dollarAmounts":[{"amount":"$14 billion","context":"multiyear contracts for Akamai Cloud collectively worth more than $14 billion"}]},"materialImpact":{"score":3,"reasoning":"Launch of two foundational AI products (Akamai Inference and Akamai Agent Sandbox) plus expanded NVIDIA Blackwell Ultra GPU fleet, deepening Akamai's AI infrastructure strategy. Material strategic milestone but a beta-stage product launch, not an immediate earnings event."},"tickerRelevance":{"others":[{"ticker":"NVDA","relevance":"supplier"}],"primary":"AKAM"},"globalImportance":45,"audienceRelevance":55,"eventTypeSecondary":["partnership"],"importanceComponents":{"tickerTier":"large-cap","eventGravity":"product launch with strategic AI pivot","sectorWeight":"high (AI infrastructure is a hot theme)","stageOfRevenue":"beta, no near-term revenue disclosure","householdBrandBoost":"moderate"}},"event_type":"product_launch","event_type_secondary":["partnership"],"sentiment":"bullish","material_impact_score":3,"narrative":"Akamai launched two new AI stack products, Akamai Inference and Akamai Agent Sandbox, bringing managed model serving and isolated agent execution to Akamai Cloud.\n\nThe launch builds on Akamai Cloud infrastructure backed by multiyear contracts worth more than $14 billion, spanning over 4,350 edge locations and an expanded GPU fleet that now includes NVIDIA Blackwell Ultra clusters, with Vera Rubin support planned for 2027.\n\nBoth products are available in beta; the move positions Akamai to capture enterprise AI inference and agentic workload revenue on a single platform, though revenue contribution remains early-stage.","key_figures":{"customDimensions":{"edge_locations":"more than 4,350","model_param_range":"8 billion to 2.8 trillion parameters","models_in_catalog":12,"gpu_memory_per_node":"2.3 TB total (288 GB HBM3e per GPU)","cloud_contracts_value":"more than $14 billion"}},"named_entities":{"people":[{"name":"Adam Karon","role":"Chief Operating Officer and GM of Akamai Cloud"},{"name":"Jon Alexander","role":"Senior Vice President of Cloud Technology"}],"products":["Akamai Inference","Akamai Agent Sandbox","Akamai Cloud","Akamai Functions","NVIDIA Nemotron 3 Ultra","OpenAI gpt-oss-120b","Google Gemma 4","NVIDIA Blackwell Ultra","NVIDIA RTX PRO 6000 Blackwell Server Edition","Spin framework"],"companies":[{"name":"Akamai","ticker":"AKAM","relationship":"filer"},{"name":"NVIDIA","relationship":"supplier"},{"name":"OpenAI","relationship":"model provider"},{"name":"Google","relationship":"model provider"}],"dollarAmounts":[{"amount":"$14 billion","context":"multiyear contracts for Akamai Cloud collectively worth more than $14 billion"}]},"model_name":"glm-5.3-flashx","prompt_hash":"sha256:5c812bd9e28b810d","schema_hash":"sha256:05005c02d9cffac9","created_at":"2026-10-27T10:30:07.705Z","global_importance":45,"audience_relevance":55,"importance_components":{"tickerTier":"large-cap","eventGravity":"product launch with strategic AI pivot","sectorWeight":"high (AI infrastructure is a hot theme)","stageOfRevenue":"beta, no near-term revenue disclosure","householdBrandBoost":"moderate"}},"durationMs":6793,"modelName":"glm-5.3-flashx"}}