EDGE by Field AI: status and timeline

Milestones

Next · Series C term sheet signed; valuation reaches $10 billion

Field AI founded, led by NASA JPL autonomy researchers2023Complete.
Series A funding round led by Khosla VenturesNov 2024Complete.
Series B funding round; revenue and signed contracts reach $100MAug 2025Complete.
EDGE robot brains deployed across 30+ customer sites globallyJun 2026Complete.
Series C term sheet signed; valuation reaches $10 billionTarget 2 Oct 2026Current milestone.
Global expansion and headcount doublingTarget 2027Not yet reached.

Upcoming

Q4 2026Series C funding round closes (next)
2 Oct 2026Series C term sheet signed; valuation reaches $10 billion
2027Global expansion and headcount doubling
2027Humanoid and manipulation capabilities launched

Current obstacles

Scaling to complex manipulationExtending the system to handle dexterous tasks requiring precise control and sensing on humanoid robots.
Cost reduction for mass productionLowering computational and energy overhead to meet manufacturing cost targets for commercial robotics.

Physics limits

On-edge compute is limitedRunning foundation models locally means constrained computational budgets compared to cloud-connected systems, trading off latency gains for capability depth.
Training data from real robots is scarceUnlike vision or language models trained on internet-scale data, robotics foundation models require expensive real-world interaction and costly hardware deployments to generate training data.

How it works

3 parts

Belief-based reasoning

World model that reasons under uncertainty

The BWM predicts what robots expect, adapting when reality differs.

On-edge deployment

Runs without cloud connectivity

EDGE operates locally on robot hardware, enabling GPS-denied deployment with minimal latency.

Physics-first foundation

Models built for robotics from scratch

FFMs are architected around physics and robotic constraints from the ground up, not adapted from vision/language models.

Spec sheet

Spec	EDGE
Customer count	30+^R (reported)
Revenue + contracts	$135M+^R (reported)
New contracts since June 2026	$35M+^R (reported)

^R reported by the company

Papers & demos

1 item

2023demo
EDGE robot brain unveiled (opens fieldai.com)Introduction of FFMs for autonomous robot operation.

About Field AI

The team behind EDGE

Field AI

EDGE robot brain, foundation models, autonomous robotics

Field AI develops Field Foundation Models for autonomous robots in real-world environments. Founded in 2023 by NASA JPL researchers led by Ali-akbar Agha-Mohammadi, it has deployed EDGE across manufacturing, construction, and industrial sectors on three continents.

Field AI has $135M+ in revenue and contracts across 30+ customers as of October 2026.
October 2026 term sheet: $700M at $10B valuation, a 5x increase from $2B in August 2025.
FFMs are purpose-built for robotics, operating on-edge without cloud dependency.

fieldai.com (opens fieldai.com)Company profile →

Founded: 20233 yrs
Headquarters: United States
Status: Private
Valuation: $10Bdeal valueOct 2026
Raised: $1.2B3 rounds
Last round: Series C (term sheet) · $700MOct 2026
Works in: RoboticsRobot brains
Coverage: 1 program · 4 updateslatest 8d agochecked 10 Oct
Lead investors: Bezos ExpeditionsKhosla VenturesGates Frontier
Customers: Caterpillar (opens caterpillar.com)Big-D Construction (opens fieldai.com)
People: Ali-akbar Agha-Mohammadi (opens fieldai.com)Founder & CEO

Spec

EDGE

Customer count

30+^R (reported)

Revenue + contracts

$135M+^R (reported)

New contracts since June 2026

$35M+^R (reported)