AI Stocks List 2026: 85+ Public AI Companies

Through AWS, Amazon is a leading AI cloud provider, offering managed models via Amazon Bedrock plus its own Trainium training chips. Anthropic and OpenAI have both made multi-year, multi-gigawatt commitments to Trainium capacity. AWS sales grew 37 percent in the second quarter of 2026, its fastest pace in 18 quarters.

Taiwan Semiconductor is the world's largest contract chipmaker and manufactures the advanced-node processors and advanced packaging used in most leading AI accelerators, including chips for NVIDIA, AMD, and hyperscaler custom silicon. High-performance computing tied to AI is now the majority of its revenue, and TSMC is expanding CoWoS packaging and overseas fabs to meet that demand.

Broadcom designs custom AI accelerator chips for hyperscalers including Anthropic, Google, Meta, and OpenAI, and supplies the high-speed Ethernet switching silicon that connects large AI clusters. AI semiconductor revenue reached 10.8 billion dollars in its fiscal second quarter of 2026, and the company guides to about 56 billion dollars for the full fiscal year. It also sells infrastructure software.

Meta runs Meta Superintelligence Labs and develops the Muse family of frontier models, including the Muse Spark reasoning models and the open-weight Muse Glimmer. AI powers ad ranking and content recommendation across Facebook, Instagram, and WhatsApp, plus the Meta AI assistant and its smart glasses. Meta is among the largest spenders on AI data centers and custom silicon.

Micron makes the high-bandwidth memory and storage that AI accelerators depend on. Its HBM4 stacks ship in volume for a lead customer's AI platform, and Micron says HBM output is fully booked through calendar 2027 with demand extending into 2028. Data center memory drives the bulk of revenue, tying results closely to AI hardware demand.

AMD designs Instinct data center GPUs and Helios rack-scale systems that compete with NVIDIA for AI training and inference, alongside EPYC server processors. Its Data Center segment supplied 58 percent of revenue in the second quarter of 2026 and more than doubled from a year earlier. Anthropic, Microsoft, OpenAI, and Oracle are among the customers deploying its AI systems.

Palantir builds data analytics and operating platforms (Gotham, Foundry) for governments and enterprises. Its AI Platform, AIP, lets organizations run large language models against their own operational data to support decisions. Revenue grew 93 percent year over year in the second quarter of 2026, led by US commercial customers, and AI software now drives the bulk of that growth.

Oracle Cloud Infrastructure has become a major AI training and inference cloud, and Oracle is a central partner in large-scale AI data-center buildouts including the Stargate project with OpenAI. Its cloud infrastructure revenue and contracted backlog grew sharply in fiscal 2026 on AI demand, and Autonomous Database and OCI AI services embed machine learning across the stack.

Arista Networks builds the high-speed Ethernet switches that connect thousands of AI accelerators inside data centers. In 2026 it introduced 1.6-terabit Etherlink platforms, including liquid-cooled options, for scale-up, scale-out, and scale-across AI fabrics. The company targets at least $3.5 billion in AI fabric revenue for 2026, and major cloud operators are among its largest customers.

Marvell designs custom AI accelerators for hyperscaler data centers and supplies the optical interconnect and switching silicon that moves data between AI chips. Data center products now account for about three quarters of its revenue. In 2026 NVIDIA invested $2 billion in Marvell and connected it to the NVLink Fusion ecosystem for rack-scale AI systems.

Vertiv provides the power management and thermal systems that AI data centers need to run dense GPU racks. Net sales grew 24 percent in the second quarter of 2026, and the company raised its full-year outlook on AI-driven demand. Management says new high-density facilities are liquid-cooled by default, and Vertiv is among a limited set of qualified hyperscale suppliers.

Nebius is an AI cloud provider that rents NVIDIA GPU capacity for training and inference workloads. It has signed multi-year capacity contracts with Microsoft and Meta and received an NVIDIA investment, and its annual recurring revenue and data-center footprint expanded through 2026. Nebius sits alongside CoreWeave as a dedicated AI compute platform.

CoreWeave is a cloud provider built specifically to rent GPU capacity for AI training and inference. Its platform is used by major AI developers, including OpenAI, Meta, and Anthropic, under multi-year capacity contracts, and its revenue backlog and contracted data-center power have grown sharply through 2026. NVIDIA holds an investment in the company.

Cerebras Systems builds wafer-scale processors and the AI supercomputers around them for training and inference workloads. It sells systems to enterprise and research customers and operates its own inference cloud service. Cerebras listed on Nasdaq in May 2026 and holds a multi-year agreement with OpenAI to supply inference compute capacity.

Super Micro Computer builds AI and high-performance server systems, including liquid-cooled rack-scale clusters for NVIDIA GPU platforms. It has announced support for NVIDIA's Vera Rubin systems and expanded its liquid-cooling manufacturing capacity. Revenue reached $39.1 billion in fiscal 2026, and the company reports a record order backlog for AI systems.

SpaceX builds rockets and operates the Starlink satellite network, and lists on Nasdaq with three reported segments (Space, Connectivity, AI). Its AI segment spans the Grok models, the X platform, and GPU data-center capacity rented to AI companies, and it produced about a third of revenue in the second quarter of 2026. Anthropic and Google contract for its data-center compute.

Tesla applies AI to autonomous driving through its Full Self-Driving software and the data from its vehicle fleet, and to robotics through the Optimus humanoid program. Its robotaxi service now operates in several US metro areas and continues to expand. Tesla designs its own AI inference chips for vehicles and robots and is building semiconductor capacity in Austin.

ASML is the sole supplier of the extreme-ultraviolet lithography machines required to manufacture leading-edge AI chips, so every advanced processor from companies such as NVIDIA and TSMC depends on its equipment. It raised full-year 2026 sales guidance as AI demand pulled forward logic and memory capacity, and Intel has qualified its High-NA systems on a production process.

Lam Research supplies wafer fabrication equipment for deposition and etch, the process steps used to build advanced logic and memory devices. Its tools are also used in high-bandwidth memory production and in advanced packaging, where chiplets and memory stacks are assembled for AI accelerators. Advanced packaging is one of its fastest growing product areas.

Applied Materials makes the equipment and process technology used to manufacture advanced semiconductors. Its AI relevance is upstream, since chipmakers building AI processors and memory rely on its deposition, etch, and inspection systems. AI infrastructure spending now drives its largest growth areas, leading-edge foundry logic, DRAM including high-bandwidth memory, and the advanced packaging used to assemble AI accelerators.

Alibaba operates e-commerce marketplaces and Alibaba Cloud, and develops the Qwen family of open-weight AI models. Cloud revenue growth accelerated on AI demand, with AI-related product revenue at a multi-billion-dollar annualized run rate, and the company is investing heavily in AI infrastructure. Its US-listed shares trade on the NYSE as American depositary shares.

Dell builds and integrates AI-optimized servers and infrastructure for enterprises and cloud operators, largely based on NVIDIA accelerators. AI server revenue and backlog both reached records in the first quarter of fiscal 2027, and Dell raised its full-year AI server outlook, making it one of the primary channels for deploying AI hardware at scale.

Arm licenses the processor architecture and IP behind most smartphones and a growing share of data center, automotive, and edge chips, increasingly tuned for AI. Data center royalties more than doubled year over year in the quarter ended June 2026. Arm also ships its own data center CPU for AI inference, launched in March 2026 with Meta as lead partner.

KLA supplies process control systems that inspect and measure semiconductor wafers during manufacturing, helping chipmakers find defects and improve yields. Its advanced packaging portfolio serves the assembly steps used for AI accelerators and high-bandwidth memory. The company sells to logic and memory manufacturers and to advanced packaging facilities.

SanDisk designs and manufactures NAND flash memory and solid state drives, and became an independent public company after separating from Western Digital in 2025. It has signed multi-year supply agreements covering AI-related storage capacity for data-center customers, tying a growing share of its enterprise flash output to AI infrastructure demand.

IBM provides enterprise AI through its watsonx platform and open Granite models, the latest being the Granite 4.1 family, paired with a large consulting practice that helps organizations deploy generative and agentic AI. Recent acquisitions of Confluent and HashiCorp strengthen the data and infrastructure foundation for running enterprise AI at scale.

CrowdStrike provides cloud-delivered cybersecurity through its Falcon platform, where AI models drive threat detection and Charlotte AI runs autonomous security agents. Its Falcon AIDR product secures enterprise AI use itself, covering prompt injection, data leakage, and unsafe agent actions across AI applications and assistants. Management positions the company as security infrastructure for AI adoption.

Amphenol manufactures connectors, cable assemblies, and interconnect systems across communications, industrial, automotive, and defense markets. Its IT datacom segment supplies the high-speed copper and optical interconnects that link accelerators, servers, and switches inside AI data centers. Data-center applications now represent a substantial share of its sales.

Seagate Technology designs and manufactures hard disk drives, including the mass-capacity nearline drives sold to cloud and data-center operators. Its HAMR-based Mozaic platform increases areal density so each drive stores more data. The company attributes current nearline demand to the data volumes generated by AI training and inference workloads.

A customer relationship management platform that embeds AI through Agentforce, deploying autonomous agents for sales, service, and marketing grounded in its Data 360 platform. Combined Agentforce and Data 360 annual recurring revenue reached about 3.4 billion dollars in the quarter ended April 2026, helped by the Informatica acquisition, and Agentforce alone passed 1 billion dollars.

Qualcomm designs the processors and on-device AI engines used in most premium smartphones and a growing range of AI PCs. It is building out a data center business around its AI200 and AI250 inference accelerators and the Dragonfly roadmap, with Meta signed for its C1000 server CPU. Its purchase of AI software company Modular adds tools for running AI workloads across chip types.

Western Digital manufactures hard disk drives for cloud and data-center storage, including the high-capacity nearline drives sold to hyperscale operators. The company separated its flash business into SanDisk in 2025 and now concentrates on hard disk technology. It ties current nearline demand to the data volumes generated by AI workloads.

ServiceNow's platform automates IT, HR, customer service, and security workflows, with AI delivered through its Otto assistant and a growing roster of AI agents. Its AI Control Tower governs AI deployed across enterprise systems, and AI annual contract value passed one billion dollars in 2026 as more customers moved agentic AI into production.

Snowflake runs a cloud data platform that lets enterprises store, query, and analyze large datasets, and build AI on top of them. Its Cortex AI services and CoWork agent (formerly Snowflake Intelligence) bring large language models to customer data, and Snowflake has deepened its Anthropic partnership to put Claude models behind those tools.

Adobe builds the Firefly family of generative AI models for images, video, and design, integrated across Creative Cloud and its GenStudio marketing suite, alongside AI assistants in Acrobat and Experience Cloud. Adobe reported that AI-first annual recurring revenue passed 500 million dollars in its fiscal second quarter of 2026, roughly triple the year-ago level.

Cloudflare operates a global network for connectivity, security, and performance, and runs AI inference at the edge through Workers AI and AI Gateway. It has built controls and payment tools that let site owners charge AI crawlers and agents for access, positioning its network as a toll point for automated traffic. The company reported that most traffic crossing its network is now non-human.

AppLovin runs a mobile advertising and marketing platform whose AXON engine uses machine learning to match and bid on ads in real time. The company opened its self-serve advertising platform to all advertisers in June 2026 and is extending it from mobile gaming into e-commerce. AI is central to how the platform targets campaigns, making it an AI-applied-to-advertising business rather than an infrastructure supplier.

Cadence makes electronic design automation software, chip IP, and system-design tools used to create advanced semiconductors. It also embeds AI in its own products through agentic design tools such as Cerebrus AI Studio and the Millennium supercomputer platform built on NVIDIA systems. Its backlog reached a record level in mid-2026 on AI and high-performance computing design activity.

Datadog provides cloud monitoring and observability software, with products that trace and troubleshoot large language models, AI agents, and the applications built on them. AI-native companies are one of its larger customer groups, and several of them spend at enterprise scale on its platform. It added new AI observability and agent-monitoring products in 2026.

Lumentum makes optical components and high-speed transceivers used to move data inside AI and cloud data centers. Cloud and AI demand drove record transceiver shipments in fiscal 2026, and the company began shipping 1.6T modules alongside optical circuit switching systems. Its lasers and photonics sit in the interconnect layer that links GPU clusters.

Synopsys makes electronic design automation software and chip IP used to design advanced semiconductors, and its Ansys acquisition added engineering simulation. AI-chip designers across the industry rely on its tools to build their processors. Synopsys is also developing autonomous agentic design workflows with Microsoft, which AMD is testing on its next-generation products.

Hewlett Packard Enterprise sells servers, storage, networking, and hybrid cloud services to enterprises, governments, and service providers. Its AI systems business builds and integrates accelerated computing clusters, including liquid-cooled designs, for customers training and running large models. The company reports cumulative AI systems orders separately from its other results.

Monolithic Power Systems designs power management semiconductors that convert and regulate voltage for electronic systems. Its enterprise data business supplies the power delivery components used on AI server boards and alongside accelerators. The company is developing higher-voltage architectures for the rack-level power designs used in newer AI systems.

Teradyne makes automated test equipment used to validate semiconductors before they ship, including the compute silicon and high-bandwidth memory that go into AI servers. The company also builds industrial and collaborative robots. Management attributes a large share of recent semiconductor test demand to artificial intelligence applications.

Coherent makes lasers, optical transceivers, and photonics components used in high-speed AI data center networks. Datacenter and communications is now its largest business and grew faster than the company overall in fiscal 2026, led by 800G and 1.6T transceivers. A March 2026 agreement with NVIDIA, which included a 2 billion dollar investment, supports its co-packaged optics roadmap.

Ciena builds optical networking systems and coherent pluggable optics that move traffic between and within data centers, along with the software that manages those networks. Cloud providers use its equipment for data-center interconnect as bandwidth requirements rise, and the company attributes a significant portion of recent demand to AI-related network traffic.

Astera Labs designs connectivity silicon (PCIe retimers, fabric switches, and smart cable modules) that links GPUs, CPUs, and memory inside AI servers and racks. Its Scorpio fabric switches entered volume production in 2026 and revenue roughly doubled from a year earlier, and the company is developing products for the UALink accelerator interconnect standard.

Credo Technology makes high-speed connectivity products, including active electrical cables, retimers, and optical DSPs, that move data efficiently inside AI data centers. Revenue more than doubled in fiscal 2026 and it guides to further revenue growth in fiscal 2027 on an optical DSP ramp. Its power-efficient interconnects are tied to AI infrastructure spending.

Celestica provides design, manufacturing, and supply-chain services for data center hardware. Its Connectivity and Cloud Solutions segment, which carries the hyperscaler and AI networking work, supplied about 80 percent of revenue in the second quarter of 2026. Growth programs include 800G and 1.6T networking along with rack-scale AI platforms tied to AMD Helios and OpenAI custom racks.

Tower Semiconductor is a specialty analog foundry that makes RF, power, sensor, and silicon-photonics chips. Its AI relevance is silicon photonics, where it fabricates the optical-engine silicon used in the high-speed interconnects that link AI data-center clusters. Photonics revenue grew sharply in 2026, and Tower holds multi-year supply contracts with customer prepayments tied to that demand.

Lattice Semiconductor makes small, low-power FPGAs used for system control, security, power management, and connectivity. Data-center AI applications and rising FPGA content per server drove its compute and communications revenue higher through 2026. In July 2026 Lattice acquired AMI, adding platform firmware and server management software to pair with its chips in cloud and AI infrastructure.

Fabrinet provides precision optical and electro-mechanical manufacturing for other companies, assembling and testing optical transceivers, modules, and photonic components. Its datacom programs include the high-speed transceivers used to connect accelerators and switches inside AI data centers. The company manufactures for networking equipment vendors and hyperscale operators.

IREN develops and operates data-center campuses and sells AI cloud services running on NVIDIA GPUs. It has signed multi-year contracts to supply GPU cloud capacity to large technology customers, including Microsoft, and is building liquid-cooled data centers at its Texas campus. It also retains a bitcoin mining business at some sites.

Tempus AI applies artificial intelligence to clinical and molecular data in oncology and other disease areas. The company sequences patient samples, builds a library of linked clinical records, and licenses de-identified data and analytical tools to pharmaceutical companies and researchers. Its diagnostics and data businesses draw on the same AI platform.

Rambus designs memory interface chips and licenses semiconductor IP used in server memory and high-bandwidth memory systems. Its DDR5 chipsets and memory controllers sit between processors and memory in data-center systems. The company has announced controller IP for HBM4E, the memory generation intended for AI accelerators and high-performance computing.

Applied Digital develops and operates purpose-built AI data centers and leases capacity to hyperscale and AI customers under long-term agreements. Its contracted leases cover multi-hundred-megawatt campuses built for high-density GPU computing. Its revenue is tied to the buildout of AI computing capacity in North America.

Pony AI develops autonomous driving systems and operates robotaxi services in Chinese cities, with stated plans to scale driverless heavy-truck fleets. Its business is built on machine-learning perception and planning software, making autonomy the core of the company rather than a feature. Its shares trade on Nasdaq as American depositary shares.

Navitas Semiconductor develops gallium-nitride and silicon-carbide power chips for smaller, more efficient power conversion. Its AI relevance is data-center power delivery, since NVIDIA selected Navitas to help develop the 800-volt architecture for high-density AI GPU racks, and the two collaborate through the NVIDIA MGX ecosystem. AI infrastructure is becoming the largest share of Navitas revenue.

SoundHound AI develops voice and conversational AI, including its Houndify speech platform and Amelia enterprise agents. In May 2026 it launched OASYS, an orchestrated agentic platform that builds and manages fleets of AI agents. Its technology powers in-car assistants, restaurant phone and drive-thru ordering, and customer service automation, and quarterly revenue grew 45 percent year over year.

Upstart runs an AI lending marketplace that uses machine-learning credit models to evaluate borrowers and match them with bank and credit-union partners for personal, auto, and home loans. Its models approve and price most loans on the platform without human review, so AI underwriting is the product rather than a feature.

Innodata prepares and engineers the data used to build and evaluate artificial intelligence models, including annotation, model evaluation, and safety testing work. Its customers include large technology companies developing generative AI systems. The company also offers platforms for agent evaluation and observability alongside its data services.

Absci is a generative AI drug discovery company. It pairs deep learning protein and antibody design with laboratory testing to create and optimize candidate medicines. Its lead AI-designed antibody, ABS-201, is in Phase 1 testing for pattern hair loss with a planned Phase 2 study in endometriosis, and the company reported positive interim safety and pharmacokinetic data in 2026.

BigBear.ai provides AI decision intelligence and analytics software, mainly for defense, intelligence, national security, and border and travel agencies. Its early 2026 acquisition of Ask Sage added a generative AI platform used across government teams, which BigBear.ai has folded into its own generative AI portfolio. Revenue and gross margin both grew in the second quarter of 2026.

One Stop Systems builds ruggedized, GPU-accelerated computers that run AI inference and sensor processing at the edge, inside vehicles, aircraft, and ships rather than in data centers. After selling its Bressner distribution subsidiary at the end of 2025, the company operates as a pure-play edge AI compute supplier, and defense, medical imaging, and autonomous equipment programs drive its recent revenue.

TSS provides rack-scale integration services for AI and high-performance computing, assembling, testing, and deploying the server racks that go into data centers. AI-driven systems integration is its main growth engine, and the company is expanding its Texas integration facility to handle higher volumes of GPU rack work.

Apple brings AI to consumers through Apple Intelligence across iPhone, iPad, Mac, and Vision Pro, combining on-device processing with its Private Cloud Compute servers. At WWDC 2026 it introduced Siri AI, a rebuilt assistant powered in part by Google Gemini foundation models and shipping with iOS 27. Apple spends far less on AI data centers than the large cloud providers.

Intel designs x86 processors for PCs and servers, including AI PC chips, and runs a contract chip-manufacturing foundry. Its data center and AI segment grew in 2026 as Xeon 6+ shipped on the company's 18A process. Intel is rebuilding its accelerator line with Crescent Island, an inference-focused data center GPU set for customer sampling in the second half of 2026.

Cisco Systems supplies networking, security, and observability products, including Ethernet switches and its own Silicon One networking chips. The company sells AI infrastructure networking to hyperscale operators and reports those orders separately from its other business. It also sells AI networking to neocloud, sovereign, and enterprise customers.

Palo Alto Networks is a cybersecurity vendor that builds AI across its platform and sells tools to secure enterprise AI use, including its Prisma AIRS product for AI models and agents. Its acquisition of CyberArk, completed in February 2026, adds identity security for human, machine, and AI agent identities. Cortex AgentiX applies autonomous agents to security operations.

Corning manufactures optical fiber, cable, and connectivity products alongside its display and specialty glass businesses. Its enterprise networks products carry traffic inside data centers, and the company has signed multi-year supply agreements with large cloud operators for that connectivity. Corning also develops multicore fiber and co-packaged optics for accelerated computing.

A global consulting and technology services firm that helps enterprises design and deploy AI across strategy, operations, and software. Accenture has reorganized around AI-led reinvention work and partners with model providers including Anthropic, OpenAI, and NVIDIA. As of fiscal 2026 it no longer reports advanced AI revenue separately, describing the technology as embedded across its services.

NXP makes processors, microcontrollers, connectivity, and security chips for automotive, industrial, and IoT markets. Its AI relevance is at the edge, combining software-defined-vehicle and driver-assistance processors with on-device neural processing from its Kinara acquisition. The company reports a growing pipeline of edge AI designs across industrial automation, robotics, and in-cabin automotive applications.

Workday provides cloud software for human resources and finance, embedding AI agents that automate recruiting, expense, and finance workflows. Its Sana platform is now the foundation for enterprise AI search, agents, and a no-code agent builder, and Workday has extended it into IT service management and travel. Thousands of customers use at least one of its agents.

GE HealthCare makes medical imaging systems, patient monitoring, and diagnostics, and embeds AI in its scanners and clinical software to speed imaging, automate measurements, and aid interpretation. It holds more AI-enabled device authorizations on the FDA's public list than any other company. Recent work includes AI auto-contouring for radiation therapy and multi-agent clinical tools built with cloud and chip partners.

Pure Storage sells all-flash data storage systems and subscription services to enterprises and cloud providers. The company licenses its DirectFlash technology to a hyperscale operator for use in custom storage infrastructure, and its systems are used for AI and analytics workloads that need high-throughput access to data.

UiPath provides automation software that combines robotic process automation with AI agents for enterprise workflows. Its agentic automation platform orchestrates AI agents, software robots, and people across business processes such as case management and document handling. AI capabilities are central to how the platform wins and expands enterprise deployments.

Aehr Test Systems makes test and burn-in equipment used to qualify semiconductors, including wafer-level burn-in systems for AI processors. Orders from AI processor customers have become a major driver of its recent business, alongside silicon carbide power devices. Its equipment screens chips for reliability before they are deployed in data centers.

Ambarella designs system-on-chip processors that run computer vision and AI inference inside devices rather than in a data center. Its chips are used in security cameras, automotive systems, and robotics, and edge AI products account for most of company revenue. The company also supplies the software tools customers use to deploy models on those chips.

PROCEPT BioRobotics makes surgical robots for urology. Its HYDROS system delivers Aquablation therapy for enlarged prostate and runs FirstAssist AI, image-recognition software that maps prostate anatomy on ultrasound and proposes a treatment plan. The FDA cleared a second-generation version of that software in 2026, so AI supports surgical planning within a robotics business.

Arteris licenses network-on-chip interconnect IP that connects the processing blocks inside complex semiconductors, including AI accelerators and automotive chips. Chip designers use its technology to move data efficiently within their silicon. The company reports growing design activity tied to AI and data-center chip development.

indie Semiconductor designs automotive mixed-signal chips for driver assistance and autonomy, including radar, LiDAR, vision processors, and connectivity. Its edge AI processors, such as the iND880 and iND881 vision and perception chips, extend that work into robotics and what the company calls physical AI. In 2026 it reported design wins at robotics makers and agreed to buy a CMOS image sensor product line.

Alpha and Omega Semiconductor makes power semiconductors, including MOSFETs, silicon-carbide and gallium-nitride devices, and power-management modules. Its AI relevance is power delivery inside AI servers, with high-current MOSFETs and a SmartClamp DrMOS line aimed at data-center racks. Management describes AI and server demand as the main source of growth in its advanced computing business.

Information Services Group is a technology research and advisory firm that now describes itself as AI-centered. It advises enterprises on AI adoption, sourcing, and governance, and publishes research on technology service providers. AI-related work accounted for more than a third of its revenue in the first half of 2026, so its exposure to the theme comes through advisory demand rather than AI software it builds.