For EdgeRunner AI, from 🤫 hussh
Private, sovereign AI for the people who serve.
EdgeRunner AI was built by a veteran, founder Tyler Saltsman, a former US Army logistics officer, for the people he served alongside: agentic AI that runs air-gapped on the device in a warfighter's hands, with no internet at all. We believe the same thing. AI should be under your control, on hardware you own. EdgeRunner serves the mission at the tactical edge. 🤫 Private Agent One serves the same people across the rest of their lives. We'd love to build for them together.
What EdgeRunner and 🤫 make possible.
You serve the warfighter with private, air-gapped, on-device AI at the classified edge.
🤫 Agent One serves the same people across their personal, financial and family life.
The whole person covered, the mission and the life behind it, by one private, sovereign philosophy.
You are an Awardable Tradewinds vendor with Space Force, Army and Air Force work.
We serve the same government and veteran communities, and Agent One is free for life to those who serve.
One clear story for the same buyer: sovereign AI for the mission and for the person.
You run on Intel, HP, AMD and NVIDIA edge silicon and are working toward IL-level accreditation.
Our edge grid runs on the same silicon, and we are building to FedRAMP High and DoD IL (in pursuit).
Shared hardware economics and a shared compliance road, walked together.
Built for the edge, from the first line of code.
EdgeRunner AI, founded in 2024 in Seattle, builds what the field needs: frontier-class AI small enough to live on a chip, fine-tuned on military doctrine, running fully air-gapped in DDIL environments (Denied, Disrupted, Intermittent, Limited).
On EdgeRunner's military benchmarks, its 20B model matches or exceeds GPT-5, with no cloud at all.
Trained on military doctrine, history, tactics and manuals. Domain-specific by design.
EdgeRunner's platform is free for Department of Defense users. An instinct we share.
A seed of about $12M led by Four Rivers, then a Series A led by Madrona with HP Tech Ventures and Alumni Ventures.
An Awardable vendor, with a Space Force contract and research CRADAs with the Army (ARL) and Air Force (AFRL).
EdgeRunner's edge is an architecture and a mission, not a feature.
Air-gapped by design
Built from the start to run completely on the device, on a workstation, laptop or smartphone, isolated from outside networks.
Domain-specific models
Compressed and fine-tuned on military data: small enough to live on a chip, expert where it matters.
Agentic at the edge
Agents that act and finish tasks for the user, in logistics, maintenance and acquisitions, with function-calling models like EdgeRunner Command.
Proven in benchmarks
Among the first to bring GPT-5-class capability fully onto the device for DDIL use.
A veteran-led mission
Founded and led by people who served. The credibility to build for the warfighter comes from having been one.
A distinguished advisory board
Retired general officers and senior technologists, including Gen. Poppas, Lt. Gen. Bender and Bob Muglia.
Where EdgeRunner is going.
Every one of these priorities has room for a private agent the person owns.
Local models as the way forward
EdgeRunner's thesis: frontier cloud AI is pricing itself into a corner, and on-device economics and control are the answer. We agree completely.
The warfighter, first
Domain-specific agents that raise the odds of the warfighter winning the fight and coming home.
Sovereign, Western open AI
A public stand for sovereign AI with a "code of honor."
A deeper DoD footprint
Growing through Tradewinds and the service research labs: Space Force, Army and Air Force.
Rigorous military benchmarks
Evaluation as a moat, in combat arms, combat medicine and logistics, so the models are trusted where it counts.
Role-specific agents
Logistics, maintenance, acquisitions and combat medicine: agents tuned to the job a service member does.
EdgeRunner 20B, EdgeRunner 2 and the field.
What EdgeRunner shipped most recently.
EdgeRunner 20B (October 2025)
A fine-tuned model that, in EdgeRunner's arXiv paper, matches or exceeds GPT-5 on four military test sets with 95%+ statistical significance, running fully air-gapped on an edge device.
EdgeRunner 2 (May 2026)
A new suite of state-of-the-art military language models, with ongoing research into on-policy distillation of private data.
Space Force and research labs
A firm-fixed-price Space Systems Command contract (April 2026) to speed acquisitions with AI, plus CRADAs with Army and Air Force research labs.
Runs on AI PCs
Deploys on everyday edge hardware from Intel, HP, AMD and NVIDIA, the same silicon our edge is built on.
Serve the same people: the mission, and the person behind it.
EdgeRunner serves the tactical, classified, air-gapped edge. 🤫 Agent One serves the same people across their personal, financial and family life. Together we cover the whole person.
Build: agents that work together on the device
EdgeRunner ships tool-calling, agentic on-device models for the mission, inside the air gap, with best-in-class compression and DDIL engineering.
🤫 Agent One is the person's own consent-first agent for the rest of life, speaking MCP, A2A, ADK, AP2 and UCP, on the same on-device patterns that power Puppy One.
A mission agent and a personal agent that cooperate by consent, never crossing the air gap, with every access receipted (PCHP), on one reference architecture for private AI.
Sell: one private-AI story
EdgeRunner is an Awardable Tradewinds vendor with Space Force, Army and Air Force work.
🤫 serves the same government, military and veteran communities, free for life to those who serve.
"Sovereign AI for the mission, and for the person behind it," one story to the same buyer.
Sell: the veteran's whole life
EdgeRunner serves the warfighter on active duty.
🤫 Agent One is free for life to every past and present service member, for their finances, data and Trust Circle.
A natural handoff at transition. The service continues into civilian life.
Buy: shared edge silicon
EdgeRunner runs on Intel, HP, AMD and NVIDIA edge hardware.
Our garage and warehouse grid runs on the same silicon.
Edge compute negotiated and bought together, with better economics for both.
Take risk: the compliance road
EdgeRunner is working toward IL-level accreditation for classified environments.
We are building to FedRAMP High and DoD Impact Levels (in pursuit).
One shared compliance path, and a public stand, written together, for private AI for those who serve.
The people building EdgeRunner AI.
Tyler Xuan Saltsman
Co-Founder and CEO
Eight years as a US Army logistics officer, including Operation Atlantic Resolve in 2017, then AWS and leading SuperCompute at Stability AI, before building the private, on-device AI the field needs. His mission: raise the odds that the warfighter wins and comes home.
Colton Malkerson
Co-Founder and COO
Former head of enterprise sales at Stability AI, with US government experience and a Stanford GSB background, building the go-to-market for sovereign, on-device AI.
The advisory board
Generals and senior technologists
Retired general officers and leaders like Bob Muglia, former CEO of Snowflake, guiding a company built to serve those who serve.
The same people, the same conviction.
EdgeRunner and 🤫 serve the same people, the men and women of the US government and armed forces, and believe the same thing: AI should be private, sovereign and run on hardware the person owns. 🤫 Agent One is free for life to every federal and state government employee and every past or present service member. EdgeRunner protects the mission at the edge. 🤫 protects the rest of the person's life.