Y Combinator F26 Batch Companies
Y Combinator F26 Companies
Startup Batches
Welcome to our interactive overview of Y Combinator's F26 batch. Below, you'll find a sortable, filterable table showcasing portfolio companies from this cohort.
Each row details:
- Company name and sector (spanning B2B SaaS, Healthcare, Fintech, Biotech, Construction, Analytics, Consumer, Deep Tech, and more)
- Funding stage to help you gauge maturity and runway
- YC Partner assignment for insight into mentorship and domain expertise
Use this table to:
- Explore Y Combinator's current investment focus, spotting which industries and technologies are gaining traction
- Identify promising directions by filtering sectors or funding stages for your own market research or strategic planning
- Quickly find founder contacts for networking, partnership, or investment outreach
This page goes live early and updates as Y Combinator announces the Fall 2026 cohort, so check back to track new companies, market trends, potential collaborators, and the next wave of disruptive startups in one centralized view.
- Company Website
- runinfra.ai
- One-Liner
- Chat-native AI model optimization platform
- Employee Count
- 1-10
- Company Website
- degla.ai
- One-Liner
- Autonomous drone fleets for emergency response
- HQ Country
- United States
- Company Website
- godhands.dev
- One-Liner
- Deterministic computer-use infrastructure for agents
- HQ Country
- United States
- Company Website
- simantic.dev
- One-Liner
- Simulated hardware for embedded QA
- HQ Country
- Canada
- Company Website
- usecollarai.com
- One-Liner
- AI deal agent for private market investors
- HQ Country
- Singapore
- Company Website
- lambdarobotics.ai
- One-Liner
- Autonomous robots for AI data centers
- HQ Country
- United States
- Company Website
- qokedas.com
- One-Liner
- Small lab teaching AI to decode physical instruments
- HQ Country
- United States
Company | Company Website | One-Liner | HQ Country | Employee Count | Founding Year | Product Description | YC Page | YC Industry | YC Partner | Functional Type | Extruct Industry | LinkedIn Profile | GitHub URL | Business Model v4 | Extruct Category | Sub-cluster | AI Pattern | Builds Hardware | Funding Verified | Delivery Model | Number of Founders | Founders LinkedIn |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Chat-native AI model optimization platform | – | 1-10 | 2025 | RunInfra is a chat-native platform for building, optimizing, and deploying open-source AI inference pipelines. It benchmarks real GPUs, applies kernel optimizations, and generates OpenAI-compatible production APIs from plain-English prompts. The company is backed by Y Combinator and presents RunInfr | B2B, Infrastructure | Nicolas Dessaigne sources: 1 | AI Research, Dev tools for AI Agents | Analytics | Developer infrastructure | DevTools | inference pipelines | AI infrastructure / tooling | Software only | – | Developer platform / API / infra | 2 | 2 founders | |||||
Autonomous drone fleets for emergency response | United States | 1-10 | 2026 | DEGLA builds autonomous aerial intelligence software for multi-drone emergency missions. Its system accepts a single natural-language command and coordinates drone fleets during search and rescue or disaster response. The homepage shows an earthquake-collapse scenario and positions the product for f | Industrials, Drones | Garry Tan sources: 1 | AI Research, Other | – | – | – | Hardware | Hardware | autonomous rescue drones | AI agent for a vertical | Hardware / physical product | – | Hardware | 1 | 1 founder | |||
Deterministic computer-use infrastructure for agents | United States | 1-10 | 2026 | GodHands provides deterministic execution infrastructure for software agents across browser and desktop environments. It describes itself as an automation API for the entire desktop, with zero-fail reliability, full audit trails, and support for modern and legacy applications. The company says it is | B2B | Nicolas Dessaigne sources: 1 | Workflow automation, AI Research | Analytics, Engineering, Infrastructure | – | – | Developer infrastructure | DevTools | deterministic execution infra | AI infrastructure / tooling | Software only | – | Developer platform / API / infra | – | – | |||
Simulated hardware for embedded QA | Canada | 1-10 | 2025 | Simantic provides simulated hardware for embedded QA and automated firmware validation. The product combines CLI access, an MCP server for AI agents, and GitHub integration for firmware testing workflows. The company says it supports mixed-signal development and mission-critical deployments. | B2B, Engineering, Product and Design | Aaron Epstein sources: 1 | Other | – | – | Developer infrastructure | DevTools | embedded QA tools | AI infrastructure / tooling | Software only | – | Developer platform / API / infra | 2 | 2 founders | ||||
AI deal agent for private market investors | Singapore | 1-10 | 2026 | Collar AI is an AI deal agent that automates the entire pre-investment process for private market investors — including deal sourcing, dataroom analysis, financial modeling, and IC memo generation. It connects to a fund's inbox and deal history to create a proprietary company brain that operates in | B2B | Grey Baker sources: 1 | AI Research, Analytics & BI: Data analytics | Asset Management | – | AI-native service | Fintech | investment deal automation | AI agent for a vertical | Software only | – | Self-serve SaaS / software | 2 | 2 founders | ||||
Autonomous robots for AI data centers | United States | 1-10 | – | Lambda Robotics builds and deploys autonomous robots for the data centers and facilities that keep AI compute running. The robots are deployed into real-world environments and improve with every deployment. The company focuses on embodied intelligence as a systems problem, integrating perception, pl | Industrials, Manufacturing and Robotics | Aaron Epstein sources: 1 | Robotics | Robotics | – | Hardware | Hardware | robotic systems | Hardware with AI | Hardware / physical product | – | Hardware | 1 | 1 founder | ||||
Small lab teaching AI to decode physical instruments | United States | 1-10 | 2026 | Qokedas converts raw signals that pass through scientific instruments into training data, enabling labs to build models that read the physical world and operate the machines that measure it. It has published the aubench-lite benchmark testing whether frontier AI agents can decode a real 1,024-microp | B2B | – | – | – | – | – | – | – | – | – | – | – | – | 3 |
Get weekly YC launch + funding round alerts
Launches, funding rounds, acquisitions, tech stack, and customers' LinkedIn and emails for every YC batch.
Delivered to your inbox before they hit the news.