AI Stocks: Public-Company Intelligence Hub
14 AI-relevant public companies, hand-curated. Each ticker links to a per-company page aggregating recent SEC filings, AI-tagged news mentions, and TensorFeed editorial. Built for human readers and AI agents doing pre-trade research. Free, sourced, no quotes or recommendations.
Silicon
(5)The dominant supplier of AI training and inference GPUs (H100, H200, B200, Rubin). Every frontier lab buys from here.
MI300 / MI325 / MI350 series Instinct accelerators; the credible second-source story for hyperscaler AI compute.
Custom AI ASICs (Google TPU partner, Meta MTIA partner), networking silicon for hyperscale fabrics.
The fab. N3 and N2 capacity backs every leading-edge AI accelerator (NVIDIA, AMD, Apple, AVGO, ARM customers).
CPU architecture underneath NVIDIA Grace, AWS Graviton, Apple silicon, and the on-device inference cohort.
Hyperscaler
(4)Azure OpenAI Service, Copilot product line, capex leader, sole exclusive OpenAI cloud partner (until the 2025 Microsoft-OpenAI reset).
Gemini frontier model line, TPU silicon, Google Cloud + DeepMind, owner of the Android Gemini Intelligence distribution lever.
AWS Bedrock + Trainium + Inferentia, $8B Anthropic investment, Project Rainier compute build for Anthropic training.
OCI GPU capacity reseller to OpenAI, xAI, and Microsoft; Stargate datacenter program partner.
AI Native
(1)Infrastructure
(1)Consumer
(3)Apple Intelligence on-device + Private Cloud Compute; iOS distribution; partner of OpenAI, Anthropic, and Google through the Extensions framework.
Llama open-weights line, Reality Labs, MSL talent group, Meta AI consumer + enterprise rollout.
FSD + Optimus humanoid robotics + Dojo training compute; xAI Colossus orbital infrastructure ties.
FAQ
What is on this page?
A curated registry of AI-relevant public companies. The cohort spans silicon (NVDA, AMD, AVGO, TSM, ARM), hyperscalers (MSFT, GOOGL, AMZN, ORCL), AI-native software (PLTR), AI infrastructure integrators (SMCI), and consumer + multi-product operators with material AI exposure (AAPL, META, TSLA). Each company has a dedicated page aggregating recent SEC filings, AI-tagged news mentions, related TensorFeed editorial, and links to primary sources. Editorial curation, redistribution-friendly data only.
Why these companies and not others?
The cohort is anchored on the AI bellwether list our SEC filings cron tracks at /api/sec/filings/recent. Inclusion criteria: publicly traded on a US exchange, AI exposure material to the equity story (not "they used Copilot once"), and a clean primary-source trail. Expansion candidates (NET, MDB, SNOW, DDOG, CRWD, CRM, VRT) sit in the queue; each addition costs Worker cron capacity, so the cohort grows deliberately.
Is this trading advice?
No. This is structured AI-sector intelligence designed for agents and humans doing their own research. Every datapoint is sourced and linked, no analyst ratings, no price targets, no recommendations. TensorFeed is data infrastructure, not a brokerage.
How fresh is the data on each company page?
SEC filings refresh every 6 hours from data.sec.gov (public domain). The news panel pulls live from /api/news on page load. The cohort registry on this index is a hand-curated static list updated when a company joins or leaves the AI bellwether list. The "last filing" timestamps shown on cards reflect the cohort snapshot at the latest build; click into a page for the live filing list.
How does this connect to TensorFeed agent payments?
The free /ai-stocks pages are the human view. Agents doing trade research can hit /api/sec/filings/recent?ticker=X (free) for filings, /api/news for headlines, and the forthcoming /api/premium/ai-companies/{ticker} (paid) for a single aggregated JSON envelope per company. The premium endpoint will be x402-priced and registered through the CDP Bazaar so any agentic-trading agent with a Robinhood Agentic Credit Card (or another budget rail) can call it directly.
Why no real-time price quotes?
NYSE and NASDAQ quote licensing makes redistribution of live price data prohibitively expensive and contractually risky. TensorFeed only ships data we can redistribute cleanly under permissive licenses (SEC public domain, MIT/CC-BY corpora, our own editorial). For live quotes, your broker is the right surface.