EveryDev.ai
Subscribe
Home
Developers

5,086+ AI companies

  • Radar
  • Trending
AI Tools by Topic
  • AI Coding Assistants
  • Agent Frameworks
  • MCP Servers
  • AI Prompt Tools
  • Vibe Coding Tools
  • AI Design Tools
  • AI Database Tools
  • AI Website Builders
  • AI Testing Tools
  • LLM Evaluations
Follow Us
  • X / Twitter
  • LinkedIn
  • Reddit
  • Discord
  • Threads
  • Bluesky
  • Mastodon
  • YouTube
  • GitHub
  • Instagram
Get Started
  • Users
  • Rate Tools
  • About
  • Editorial Standards
  • Corrections & Disclosures
  • Community Guidelines
  • Advertise
  • Contact Us
  • Newsletter
  • Submit a Tool
  • Start a Discussion
  • Write A Blog
  • Share A Build
  • Terms of Service
  • Privacy Policy
Explore with AI
  • ChatGPT
  • Gemini
  • Claude
  • Grok
  • Perplexity
Agent Experience
  • llms.txt
Theme
With AI, Everyone is a Dev. EveryDev.ai ©
    1. Home
    2. Developers
    3. Runta

    Runta

    Runta is the execution layer for AI agents: a secure runtime in which agents can run computers, use credentials without seeing them, and operate under governed network and resource policies. It aims to make production agents safer, more controllable, recoverable, and less expensive to operate.

    Visit Website

    At a Glance

    1Tool Listed
    6Products
    11Capabilities
    Discussions
    San Francisco, CaliforniaHeadquarters
    2025Est.
    6Employees
    $20MRaised
    Focus Areas
    AI Infrastructure
    Compute Optimization
    Agent Harness
    Connect
    Latest News
    Introducing FrontierHarness Eval v1.1Nov 17, 2026
    How Runta protects your secretsNov 14, 2026
    Markets
    • Enterprises deploying autonomous AI agents
    • AI product companies building agent applications
    • Engineering and software-development teams using coding agents
    • Security-conscious organizations requiring credential isolation and governed network access
    • +2 more

    AI Tools by Runta

    (1)
    View Runta
    Runta tool icon

    Runta

    Managed Runtime for AI Agents

    Agent HarnessAI InfrastructureCompute Opt.

    Discussions

    No discussions yet

    Be the first to start a discussion about Runta

    Latest News

    11/17/2026

    Introducing FrontierHarness Eval v1.1

    runta.com
    11/14/2026

    How Runta protects your secrets

    runta.com
    10/01/2026

    Grok Bot vs Muse: What Happens After Prompt Injection?

    runta.com
    09/22/2026

    How We Built Errand with One Developer in One Week

    runta.com

    Products & Services

    6
    Runta Runtime
    July 16, 2026

    Managed, persistent and isolated computers for AI agents, with OS-level execution controls, governed egress, credential injection, checkpoints, auto-suspend/resume, and memory auto-scaling.

    Cloud Agents

    A managed-agent API and runtime experience providing persistent computers, conversations, streaming, runs, events, artifacts, computer sessions, and resumable agent work.

    Errand
    September 17, 2026

    An Apache-2.0 open-source AI teammate with its own persistent cloud computer. It lets users hand off tasks, close the app, and return to the same conversation, machine, files, and results.

    Exo Harness for Runta Runtimes
    September 9, 2026

    A built-in Exo runtime that separates a durable Harness from replaceable agent Executors and an isolated Sandbox, while Runta supplies protected model credentials, checkpoints, lifecycle management, auto-scaling, and network access control.

    Market Position

    Runta positions itself as an execution layer beneath the agent rather than an agent model or a front-end framework. It explicitly compares the control model with cloud-computer agent products such as Grok Bot and Muse, emphasizing enforcement at the runtime and egress boundary: isolated computers, policy-controlled destinations, credential injection without exposing secrets, persistent state, and auditability. Its broader competitive set includes managed agent sandboxes and runtime platforms, while Runta differentiates around OS/network-level governance, token-cost controls, and recoverable execution.

    Leadership

    Founders

    GD

    Guanlan Dai

    Founder and CEO. Dai was an early engineer and technical lead on Cloudflare's Edge team, where he worked on the edge proxy, CDN, caching, WAF, and front-line request path. He later became a founding engineering leader at Kong, building teams and products behind Kong Gateway, Kong Cloud, the Kubernetes Ingress Controller, and Kong's AI infrastructure products.

    Executive Team

    GD

    Guanlan Dai

    Founder and CEO

    Former early Cloudflare Edge engineer and Edge Platform leader; later a founding engineering leader at Kong responsible for Kong Gateway, Kong Cloud, Kubernetes Ingress Controller, and AI infrastructure products.

    Founding Story

    Runta was started after Guanlan Dai's experience building Cloudflare's and Kong's boundary infrastructure exposed a limitation of conventional security controls: software generally decides its behavior before execution, while AI agents make decisions dynamically as they run. Runta's initial vision was therefore to put governance underneath the agent, at the execution boundary where filesystem, process, credential, network, token, and compute actions become real.

    Business Model

    Revenue Model

    Prepaid, usage-based billing for runtime resources; customers pay while agents are running and for memory/disk during paused or suspended states. There is no subscription or monthly minimum, and automatic top-ups are optional.

    Pricing Tiers

    Usage-based Runta runtime
    vCPU $0.0504/vCPU-hour; memory $0.0162/GiB-hour; disk $0.000108/GiB-hour

    Running state bills vCPU, memory, and disk; paused state bills memory and disk; suspended/shut-down state bills disk only; deleted runtimes incur no charge. New accounts receive $50 in free credit with no credit card required.

    Target Markets

    Industries & Segments
    • Enterprises deploying autonomous AI agents
    • AI product companies building agent applications
    • Engineering and software-development teams using coding agents
    • Security-conscious organizations requiring credential isolation and governed network access
    • Teams operating long-running, asynchronous, browser, or computer-use agents
    • Researchers and engineering teams evaluating agent harnesses
    Use Cases
    • Running coding agents and autonomous agents in persistent cloud computers
    • Enterprise deployment of agents that need controlled access to internal services and credentials
    • Long-running asynchronous tasks that continue after a user closes the app
    • Secure browser and computer-use automation
    • Building agent products without implementing runtime provisioning, isolation, lifecycle, credential security, and durable run state
    • Benchmarking and evaluating coding-agent harnesses under consistent environments

    Quick Facts

    Headquarters
    San Francisco, California, United States
    Founded
    2025
    Employees
    6
    Total Funding
    $20 million
    Investors
    Andreessen Horowitz (a16z), Google DeepMind / Jeff Dean
    Office Locations
    San Francisco

    Funding History

    Seed$20 million
    July 16, 2026
    More than $100 million post-money valuation
    Andreessen Horowitz (a16z), led by Martin Casado

    History & Milestones

    November 17, 2026

    Released FrontierHarness Eval v1.1, expanding the benchmark to 22 harnesses, 5 models, 62 tasks, 4 attempts, and 27,280 runs.

    September 22, 2026

    Published the build story for Errand, describing how one developer built the desktop app in about a week on top of Runta's runtime and Cloud Agents APIs.

    September 17, 2026

    Introduced Errand, an open-source alternative to Grok Bot and Muse that uses a persistent Runta computer for asynchronous agent work.

    September 9, 2026

    Introduced Exo Harness as a built-in runtime for Runta, combining Exo's agent architecture with Runta-managed state, secrets, checkpoints, scaling, and network controls.

    September 1, 2026

    Introduced FrontierHarness Eval, a benchmark of production coding-agent harnesses running in standardized Runta environments.

    Key Capabilities

    11
    Prompt and tool-output compression to reduce token usage and spend
    Token X-ray and network-layer visibility into model-call usage
    Isolated agent virtual machines and persistent filesystems
    Credential stubs and just-in-time secret injection at the egress gateway, keeping real keys out of prompts and runtimes
    Allowlisted and governed network egress with destination controls
    OS/process, filesystem, network, credential, and task-policy recording for audit and debugging

    Integrations & Partnerships

    Platform Integrations

    • CLI for macOS and Linux
    • Python SDK (runta-sdk)
    • TypeScript SDK (@runta/runta-sdk)
    • Agent presets and workflows for Codex CLI, Claude Code, OpenClaw, and Pi
    • Exo Harness runtime integration
    • Model-provider endpoints configured through Runta's secret and egress controls
    • SSH-capable agents and web-terminal access
    • Errand desktop client and GitHub open-source repository

    Key Partnerships

    Baseten: Runta partnered with Baseten to run the FrontierHarness Eval v1.1 benchmark using Baseten-served models.

    Connect

    Website
    runta.com
    GitHub
    runta-dev
    X / Twitter
    runta
    LinkedIn
    runta-inc
    Discord
    62d4bkaTnS

    AI Topics

    3

    Runta focuses on these topics:

    AI Infrastructure(1)
    Compute Optimization(1)
    Agent Harness(1)
    Back to all developersSuggest an edit