Connect your agent to akta.pro and build company lists like this across sectors and geos.
Private US AI infrastructure startups companies, Series A or later, ranked by Q3 2026 primary-subject news volume: funding, products and services, revenue model, news signals and headcount growth for each.
| Company | Funding | Products & Services | Revenue Model | Signals (Q3 2026) | Headcount growth (Q3) |
|---|---|---|---|---|---|
| Databricks ↗ Data & AI Platform 🇺🇸 San Francisco, USA | Funding $41.61B Last raised $5.0B Aug 2026 · Private Equity | Unified lakehouse platform combining data engineering, analytics, ML and AI agents built on Apache Spark and Delta Lake, serving 20,000+ customers at a $6.9B revenue run-rate. | Subscription Recurring Usage Based Freemium Marketplace Commission Professional Services | 650 in Q3 View signals (149) → | +3% in Q3 13,816 → 14,167 employees |
| Together AI ↗ AI Cloud Infrastructure 🇺🇸 San Francisco, USA | Funding $1.33B Last raised $800.0M Jul 2026 · Series C | AI Native Cloud delivering GPU compute, serverless and dedicated inference and fine-tuning for 200+ open-source models with proprietary kernel-level optimizations. | Usage Based Subscription Recurring Managed Services | 176 in Q3 View signals (79) → | +6% in Q3 421 → 445 employees |
| VAST Data ↗ AI Data Storage 🇺🇸 New York, USA | Funding $1.98B Last raised $1.00B Apr 2026 · Series F | VAST AI Operating System unifying storage, database and compute on DASE architecture, delivering exabyte-scale, millisecond-latency data infrastructure for AI factories. | Subscription Recurring Usage Based Hardware Sales Professional Services Managed Services One Time License | 111 in Q3 View signals (55) → | +3% in Q3 1,246 → 1,281 employees |
| Lambda ↗ GPU Cloud Infrastructure 🇺🇸 San Francisco, USA | Funding $6.12B Last raised $1.01B Oct 2026 · Debt-Funded | Single-tenant NVIDIA GPU clusters (Superclusters, 1-Click Clusters, Instances) for AI training and inference, serving hyperscalers, frontier labs and federal agencies. | Usage Based Subscription Recurring Managed Services Hardware Sales | 109 in Q3 View signals (45) → | +6% in Q3 787 → 831 employees |
| Fluidstack ↗ AI Cloud Infrastructure 🇺🇸 New York, USA | Funding $3.78B Last raised $1.50B Sep 2026 · Private equity round | Gigawatt-scale AI compute campuses built via a vertically integrated power-to-console stack, under a multi-billion-dollar Anthropic build-out among other frontier labs. | Subscription Recurring Usage Based Managed Services | 72 in Q3 View signals (44) → | +70% in Q3 358 → 608 employees |
| Modal Labs ↗ Cloud GPU Infrastructure 🇺🇸 New York, USA | Funding $465.5M Last raised $355.0M May 2026 · Series C | Serverless GPU cloud letting developers run inference, training and sandboxed code execution via a Python SDK, scaling from 0 to 1,000+ GPUs with sub-second cold starts. | Usage Based Freemium | 58 in Q3 View signals (26) → | +19% in Q3 216 → 256 employees |
| LangChain ↗ AI Agent Development Platform 🇺🇸 San Francisco, USA | Funding $160.0M Last raised $125.0M Oct 2025 · Series B | Open-source agent frameworks (LangChain, LangGraph, Deep Agents) plus the commercial LangSmith platform for building, evaluating and monitoring production AI agents. | Subscription Recurring Usage Based Freemium Professional Services | 58 in Q3 View signals (48) → | +17% in Q3 400 → 469 employees |
| Scale ↗ AI Data Infrastructure 🇺🇸 San Francisco, USA | Funding $15.90B Last raised $14.30B Jun 2025 · Corporate-Funded | End-to-end AI data platform spanning human-in-the-loop labeling, GenAI fine-tuning, evaluation and government AI deployment via Scale Donovan. | Subscription Recurring Managed Services Usage Based | 54 in Q3 View signals (33) → | +3% in Q3 7,249 → 7,485 employees |
| TensorWave ↗ AI Cloud Infrastructure 🇺🇸 Las Vegas, USA | Funding $496.7M Last raised $350.0M Jun 2026 · Series B | AMD-exclusive AI cloud platform on Instinct accelerators with ROCm software, managed orchestration and liquid cooling for AI labs and enterprises. | Usage Based | 54 in Q3 View signals (27) → | +12% in Q3 163 → 182 employees |
| Ollama ↗ AI Developer Tools 🇺🇸 Palo Alto, USA | Funding $80.1M Last raised $65.0M Jul 2026 · Series B | Open-source local inference runtime letting developers run LLMs on their own hardware, plus Ollama Cloud for managed datacenter inference of larger models. | Subscription Recurring Freemium | 48 in Q3 View signals (28) → | n/d |
| Gimlet Labs ↗ AI Inference Infrastructure 🇺🇸 San Francisco, USA | Funding $392.0M Last raised $300.0M Sep 2026 · Series B | Serverless multi-silicon inference cloud disaggregating agentic workloads across CPUs, GPUs and SRAM-centric accelerators for 3 to 10x performance gains. | Usage Based Subscription Recurring | 41 in Q3 View signals (10) → | +33% in Q3 54 → 72 employees |
| Prime Intellect ↗ AI Training Infrastructure 🇺🇸 San Francisco, USA | Funding $200.4M Last raised $130.0M Jul 2026 · Series A | Open-stack platform aggregating global GPU compute with 2,500+ reinforcement-learning environments, hosted post-training, evaluations and one-click inference for AI agents. | Usage Based Subscription Recurring Marketplace Commission | 40 in Q3 View signals (20) → | +16% in Q3 70 → 81 employees |
| Baseten ↗ AI Inference Infrastructure 🇺🇸 San Francisco, USA | Funding $2.39B Last raised $1.50B Jun 2026 · Series F | Full-stack inference platform with custom model runtimes and kernels, offering dedicated deployments, model APIs and a white-label gateway across 18+ cloud providers. | Usage Based Subscription Recurring | 39 in Q3 View signals (20) → | +28% in Q3 379 → 487 employees |
| GMI Cloud ↗ GPU Cloud Infrastructure 🇺🇸 Mountain View, USA | Funding $750.0M Last raised $223.0M Sep 2026 · Series B | AI-native GPU cloud with bare metal NVIDIA compute, a 200+ model inference API (MaaS/AgentBox) and sovereign AI Factory deployments for enterprises and governments. | Usage Based Subscription Recurring Transaction Fee | 31 in Q3 View signals (14) → | +4% in Q3 148 → 154 employees |
| Bespoke Labs ↗ AI Development Infrastructure 🇺🇸 Mountain View, USA | Funding $40.0M Last raised $31.8M Jul 2026 · Series A | RL environments and synthetic data curation infrastructure, including the Bespoke Curator library and MiniCheck hallucination-detection model for AI training. | Usage Based Freemium | 18 in Q3 View signals (9) → | -15% in Q3 72 → 61 employees |
| Domino Data Lab ↗ Enterprise MLOps 🇺🇸 San Francisco, USA | Funding $223.6M Last raised Undisclosed Aug 2025 · Series Unknown | Enterprise AI platform unifying model development, MLOps, governance and agentic AI for regulated industries, used by 20%+ of the Fortune 100. | Subscription Recurring Professional Services | 16 in Q3 View signals (9) → | +3% in Q3 265 → 273 employees |
| Runpod ↗ Cloud GPU Infrastructure 🇺🇸 San Francisco, USA | Funding $122.0M Last raised $100.0M Jun 2026 · Series A | On-demand GPU compute, serverless inference and multi-node clusters across 31 regions and 30+ GPU SKUs, serving over one million developers self-serve. | Usage Based Subscription Recurring Marketplace Commission Transaction Fee Managed Services | 16 in Q3 View signals (12) → | +21% in Q3 118 → 143 employees |
| PaleBlueDot AI ↗ AI Compute Platform 🇺🇸 Palo Alto, USA | Funding $615.0M Last raised $200.0M Oct 2026 · Series C | PBD TokenRouter unifies access to frontier AI providers with smart routing and failover, plus Token Factory delivering reserved NVIDIA B300 GPU clusters. | Usage Based Subscription Recurring | 14 in Q3 View signals (7) → | +10% in Q3 30 → 33 employees |
| DataDirect Networks ↗ AI Storage Infrastructure 🇺🇸 Chatsworth, USA | Funding $309.9M Last raised $300.0M Jan 2025 · Private equity round | AI-native storage appliances (EXAScaler, Infinia, AI400X) and the Horizon orchestration layer eliminating GPU idle time for hyperscalers and sovereign AI programs. | Hardware Sales Licensing Royalties Subscription Recurring Usage Based Professional Services | 14 in Q3 View signals (9) → | +6% in Q3 1,620 → 1,723 employees |
| Superb AI ↗ Computer Vision MLOps Platform 🇺🇸 San Mateo, USA | Funding $47.9M Last raised $401,593 May 2026 · Grant | End-to-end computer vision MLOps platform with the ZERO Vision Foundation Model for zero-shot detection, serving 100+ enterprises in autonomous systems and manufacturing. | Usage Based Subscription Recurring Professional Services Marketplace Commission | 14 in Q3 View signals (8) → | -2% in Q3 65 → 64 employees |
| Exa ↗ AI Search Infrastructure 🇺🇸 San Francisco, USA | Funding $357.0M Last raised $250.0M May 2026 · Series C | AI-native web search platform with proprietary crawlers, embeddings and vector databases, purpose-built as a unified API for AI agents and coding assistants. | Usage Based Subscription Recurring Affiliate Referral | 13 in Q3 View signals (13) → | +20% in Q3 203 → 243 employees |
| Runware ↗ AI Inference Infrastructure 🇺🇸 San Francisco, USA | Funding $69.0M Last raised $50.0M Dec 2025 · Series A | Unified API aggregating 400,000+ AI models across image, video, audio and LLM modalities, powered by the proprietary Sonic Inference Engine on custom hardware. | Usage Based Subscription Recurring | 13 in Q3 View signals (6) → | +14% in Q3 63 → 72 employees |
| Thunder Compute ↗ GPU Cloud Infrastructure 🇺🇸 San Francisco, USA | Funding $37.9M Last raised $13.0M Aug 2026 · Series A | On-demand GPU cloud offering NVIDIA RTX A6000, A100 and H100 instances at 80% below AWS, using proprietary GPU-over-TCP virtualization and IDE integrations. | Usage Based | 13 in Q3 View signals (6) → | +8% in Q3 13 → 14 employees |
| Parallel ↗ AI Web Research APIs 🇺🇸 San Francisco, USA | Funding $260.0M Last raised $100.0M Apr 2026 · Series B | Programmatic web API platform letting AI agents search, extract and monitor the open web via a proprietary index and citation-backed agentic APIs. | Usage Based Subscription Recurring Transaction Fee Marketplace Commission Freemium | 12 in Q3 View signals (7) → | +22% in Q3 119 → 145 employees |
| Mem0 ↗ AI Memory Infrastructure 🇺🇸 San Francisco, USA | Funding $48.3M Last raised $20.0M Oct 2025 · Series A | Managed memory layer adding persistent, self-improving context to LLM apps and agents via a hybrid vector-graph-key-value datastore, used by 90,000+ developers. | Freemium Subscription Recurring Usage Based | 8 in Q3 View signals (8) → | +8% in Q3 24 → 26 employees |
An AI infrastructure startup sells the compute, software and data layers that other companies use to train, serve and run AI models. The category is also called the AI stack, AI compute or, for GPU clouds, neoclouds.
Compute providers such as Lambda, Fluidstack, TensorWave, GMI Cloud, Runpod and Thunder Compute rent GPU capacity. Inference and serving platforms such as Together AI, Baseten, Modal Labs, Gimlet Labs, Runware and Deep Infra turn that capacity into model APIs.
Data and storage players such as VAST Data, DataDirect Networks, Databricks and Scale supply the data pipelines models depend on. Developer and agent tooling such as LangChain, Ollama, Mem0, Exa and Parallel sits closest to application builders.
The category is defined against the hyperscalers, AWS, Microsoft Azure and Google Cloud, which sell the same layers inside broad cloud platforms. It also excludes chip designers, which supply the silicon rather than the service.
Pricing in the table follows consumption: 23 of 25 companies charge at least partly on usage, by GPU hour, token or API call, and 19 of those layer it with subscriptions or managed services.
Companies are tagged against standard industry codes and against akta.pro's internal taxonomy, so the same set can be rebuilt by code, by segment, or by a combination of both.
NAICSIndustry codes are provided for cross-referencing against public filings and procurement records. akta.pro classification is independent of them, so a company may carry an internal segment without a matching NAICS or SIC registration.