Foxglove announced a $40 million Series B on November 11, 2025, led by Bessemer Venture Partners, to expand its software for recording, searching, visualizing and operating on robotics data. The round backs infrastructure for autonomous machines—not a new robot, foundation model or hardware platform.
What Foxglove announced
Foxglove’s first-party announcement says the company raised $40 million in Series B financing on November 11, 2025. Bessemer Venture Partners led the round. Eclipse Capital, Amplify Partners and Icehouse Ventures also participated, along with angel investors Tobi Lütke, Alex Kendall, Milan Kovac, Brad Porter, Boris Sofman, Kevin Peterson, Chris Walti, Robert Sun and Lindon Gao.
| # | Preview | Product | Price | |
|---|---|---|---|---|
| 1 |
|
Nimo AI NAS, Agentic Computer Mini PC and AI Server, AMD Ryzen 7 PRO 8845HS(up to 5.1 GHZ, beat... | $1,999.99 | Buy on Amazon |
A Business Wire release dated November 12, 2025 names Bessemer, Eclipse and Amplify but not the full list published by Foxglove. Neither announcement provides a valuation, revenue, profitability figure or independently verified cumulative funding total.
Foxglove describes the financing as support for the “future of Physical AI.” In practical terms, the company is investing in the software layer that helps robotics teams turn machine-generated data into better software, models and operations.
#1 Best Overall
- [Local AI Inference & 70B Model Ready] Equipped with the AMD Ryzen 7 PRO 8845HS processor, NEXUS is engineered for heavy local AI workloads. With a full-size GPU bay, it runs 70B LLMs natively without an internet connection. Ideal for AI developers and tech enthusiasts who need private environment for coding and model testing.
- [132TB Mass Storage with ZFS Integrity] Features a hybrid storage architecture (3×NVMe + 4×3.5" HDD) supporting up to 132TB. Utilizing the enterprise-grade ZFS file system and ECC memory, it prevents data corruption and bit rot—a must-have for professional photographers and video editors safeguarding 4K/8K RAW footage.
- [OpenClaw-Driven Automation Workflow] The built-in OpenClaw execution layer allows complex automated tasks to be processed locally. Even when offline, your backup schedules and AI file organization continue seamlessly. Say goodbye to monthly cloud subscriptions and high latency.
- [Dual 10GbE & USB4 Ultra-Connectivity] Experience server-class speeds with dual 10GbE ports and a 40Gbps USB4 interface. It enables multi-user real-time collaboration on large project files directly from the NAS, ensuring zero-lag editing for creative studios and production teams.
- [Open-Source ZimaOS for Total Privacy] Running on the fully open-source ZimaOS, NEXUS ensures your data stays physically on-premise with no backdoors. It acts as a "Digital Fortress" for privacy-conscious families and small businesses who demand absolute data sovereignty.
What Foxglove sells
Founded in 2021 by people with backgrounds including Cruise, Aurora, Amazon Robotics, Stripe and Coinbase, Foxglove describes itself as a data and observability platform for Physical AI. Its product materials cover several connected parts of a robotics workflow:
- Visualization: synchronized views of video, 3D scenes, audio, GNSS, sensor readings and telemetry.
- Data management: storage, indexing, search and querying for recorded robot sessions.
- Replay and debugging: engineers can review what a robot perceived and did during a failure rather than relying only on conventional application logs.
- Fleet and device workflows: tools for moving from individual experiments to operational deployments.
- Integrations and extensions: support for existing robotics systems, formats and developer workflows.
The company’s open-source MCAP format is a central part of that stack. Foxglove says MCAP launched in 2022 and is included by default with ROS 2 and NVIDIA Isaac frameworks. MCAP is intended as an interoperable recording format for multimodal robotics data; it does not remove the need to define message schemas, maintain accurate timestamps, synchronize clocks, choose storage systems or manage permissions.
More information about the format is available at mcap.dev.
Why robotics data needs specialized infrastructure
A robot’s “log” is usually a synchronized record of many streams: cameras, lidar or other 3D sensors, microphones, GPS or GNSS, joint and vehicle state, control commands and software events. Those streams can be large, arrive at different rates and be generated at the edge, where network bandwidth is limited.
When a vehicle makes a wrong turn or a warehouse arm drops an object, the useful question is often temporal: what did each sensor show at the same moment, what did the autonomy stack infer, and which action followed? Teams may need to replay thousands of runs to find similar conditions, compare a software change with previous behavior, or select data for evaluation and model improvement.
Foxglove says production fleets can generate data at petabyte scale. That is a description of the problem the company targets, not evidence that every customer stores petabytes. Large volumes also create costs and governance obligations for retention, indexing, egress, access control and compliance.
The Physical AI thesis
Foxglove uses “Physical AI” narrowly for systems that sense the physical world, process multimodal information, make decisions and act through robots or autonomous machines. The company points to manufacturing, logistics, transportation, agriculture, construction, aerospace, defense, automotive, drones, marine systems and consumer robotics.
The investment thesis is a data flywheel:
- Sensors and onboard software produce recordings.
- Data is transferred from edge environments and organized.
- Engineers search for failures or unusual events and replay them.
- Teams curate, label, evaluate or otherwise prepare data for software and model improvements.
- Updated systems are deployed and monitored.
- New operational data feeds the next development cycle.
That makes Foxglove an infrastructure provider beneath autonomous machines. The round does not show that robotics deployment, safety certification or unit economics have been solved.
Customers and reported adoption
Foxglove’s announcements name or reference NVIDIA, Amazon, Anduril, Wayve and Dexterity. Its own Series B post also names Wayve, Waabi, Saronic, Bedrock Robotics and The Bot Company. These are companies Foxglove says use or trust its platform; the announcements do not establish contract size, deployment scope, exclusivity or customer satisfaction.
Foxglove also reports tens of thousands of developers and hundreds of customers. Those figures are company-reported. A customer logo or developer count alone cannot show how many teams run the product in production or how much revenue each account generates.
How Foxglove says it will use the money
Foxglove says the financing will fund:
- Deeper visualization capabilities.
- Expanded data-management functionality.
- Support for the complete data lifecycle, from early prototypes through production-scale deployments.
- Hiring in machine-learning platforms, data infrastructure, dataset curation, evaluation and validation, and visualization.
“Complete data lifecycle” should be read as Foxglove’s product ambition, not as a claim that one platform replaces every robotics system. Teams may still need separate products for robot control, simulation, annotation, dataset governance, model training, CI/CD, safety certification, incident response and long-term archival.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What the financing means for the Physical AI market
The round is evidence of investor conviction that robotics companies will need specialized developer and data infrastructure as they move beyond demonstrations into factories, warehouses, roads, farms and other difficult environments. Bessemer’s description of Foxglove as a category leader is an investor opinion, not an independently established market ranking.
Free tools Windows power users keep installed
One-click scans. No signup required.
The more durable question is whether infrastructure software can become a repeatable budget item across many robotics businesses. That depends on integration effort, data volumes, deployment requirements and whether teams choose a commercial platform instead of extending open-source tools and internal data lakes.
Pricing and buyer considerations
Foxglove’s pricing page, viewed August 18, 2026, lists the following starting points:
| Plan | Published terms | Best understood as |
|---|---|---|
| Free | $0 per month; 10 GB storage, three visualization users, five devices and one project | Individual experimentation and small projects |
| Pro | $20 per month plus usage; one TB storage, three developer seats and five devices included | Professional teams, with additional usage charges |
| Enterprise | Custom pricing; self-hosted data and forward-deployed engineering listed as available | Security-sensitive, large or customized deployments |
| Academic | Free for qualifying .edu or .ac users | Eligible academic users |
On Pro, the pricing page lists $42 per additional developer seat per month and $20 per additional connected device per month. Storage, queries, indexing and bandwidth can also affect the bill. Foxglove references bring-your-own storage for AWS S3, Google Cloud Storage and Azure Blob Storage, but retaining control of storage does not automatically supply robotics-specific indexing or visualization.
Before a procurement decision, a team should model retained versus deleted data, query hours, indexed terabytes, network transfer, active devices and required seats. It should also confirm whether its chosen plan includes the needed data residency, SSO, audit controls, support commitments, self-hosting or air-gapped operation. See the current Foxglove pricing page for terms that may change.
Recommended Free Tools
Who should evaluate it—and who may not need it?
Likely fit
- Robotics or autonomy teams using ROS, ROS 2, MCAP, video, 3D, sensor and telemetry data.
- Startups moving from prototype recordings to field or fleet deployments.
- Organizations that need shared replay, search and debugging instead of one engineer’s local tools.
- Teams considering hosted, on-premises, bring-your-own-storage or air-gapped options.
Possible poor fit
- Hobbyists with a few local log files and no collaboration requirement.
- Organizations with a mature internal data lake, observability system and robotics visualization stack.
- Projects needing a complete simulation, annotation, model-training or robot-control platform.
- Very large fleets whose usage-based storage, indexing, query, bandwidth, seat and device costs exceed the cost of internal tooling.
An open-source MCAP and local visualization stack can reduce licensing expense while shifting integration, maintenance and support work to the team. Existing ROS tooling or cloud object storage may already cover part of a workflow, but teams should compare the engineering required for cross-team search, synchronized replay and fleet operations.
What remains unknown
The announcements do not disclose Foxglove’s revenue mix between hosted and self-hosted deployments, paid-customer count, production usage rate, customer-level spending or valuation. They also do not establish how the platform compares economically with internal systems at very large fleet sizes. Those are important questions for investors and buyers, and the financing announcement alone cannot answer them.
Foxglove’s Series B is therefore best understood as a bet on the data layer beneath Physical AI. It gives the company resources to expand visualization, data management and lifecycle tooling, while leaving the harder market questions—deployment reliability, safety, governance and sustainable unit economics—to be proven in operation.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errors
