AgentSky is a cloud platform for running AI agents, operated by HeyMall, Inc. under the AgentSky name. Its media kit lists Xiaoyin Qu and Darren Lu as founders, and the company states that it operates from the United States. Developers choose a harness and a model, launch it in an isolated cloud environment, and reach the same persistent session from the web, the REST API, or the sky CLI, with AgentSky managing the runtime, workspace, tools, and event stream.
The platform exposes eight harnesses behind one API contract: Claude Code, Codex, Hermes, OpenClaw, pi, DeepSeek Harness, Kimi Code, and opencode, paired with Claude, GPT, Gemini, DeepSeek, Kimi, GLM, and Grok models. Sessions retain state between messages and suspend automatically when idle. Add-ons attach to the same call, covering web search, browser control, image and video generation, transcription, app connectors such as Gmail, Slack, GitHub, Notion, and Google Drive, and channels including Slack, Telegram, Discord, WhatsApp, and iMessage. The company also publishes an Agent Arena leaderboard that scores harness and model pairs on Elo, median cost, and speed from head-to-head runs on real tasks.
Billing is usage-based rather than subscription: prepaid credits are drawn on two meters, model tokens at published provider list prices with no markup, and compute charged per second while the machine is awake, from about 0.021 USD per hour for a 1 GB machine to 0.071 USD per hour at the 4 GB default. Suspended and parked agents are not billed for compute, and new accounts start with 3 USD in credit. The site carries a Product Hunt number one Product of the Day badge and publishes customer case studies from Tycoon, HeyBoss AI, and WebJourney, which report customer-provided figures of 150,000, 250,000, and 50,000 or more agents launched through the API. AgentSky does not publish its founding year, headcount, or funding history.
All products by AgentSky reviewed by our AI panel of experts.
Six independent reviewer personalities assessed AgentSky's product on their own terms. They landed within 0.9 points of each other, an unusually strong consensus.
Averaged across 1 reviewed product in our catalog. Scores are out of 10.
Who AgentSky competes with, and how they differ.
Warp's Oz platform is the closest functional match: it also runs Claude Code, Codex and its own agent inside sandboxed cloud containers across any model, driven from CLI or API. It sells to enterprise engineering teams as a per-seat subscription with concurrency caps (free/4 concurrent up to $180/mo/40 concurrent) wrapped in audit trails and access control, where AgentSky meters per-second compute plus pass-through model tokens on one key with no seat commitment.
1 product reviewed on TopReviewedOpenRouter is the model router AgentSky names itself after ("the OpenRouter for agents"): one key across hundreds of models at provider list prices with no markup on tokens. It stops at the inference call, with no agent harnesses, no persistent machines that suspend and wake, and no chat-channel surface, so the caller still supplies the runtime AgentSky sells.
1 product reviewed on TopReviewedVercel's AI SDK HarnessAgent API plus Vercel Sandbox and AI Gateway cover the same "one interface, many harnesses" idea (Claude Code, Codex, Pi today), but ship as an SDK the developer wires into their own application and deployment, billed as Vercel platform usage. AgentSky exposes the equivalent as a hosted endpoint where session lifecycle, suspension and eight harnesses are managed server-side.
2 products reviewed on TopReviewedAnthropic's Claude Managed Agents host an agent server-side with a managed sandbox and a filesystem that persists across interactions, the same shape as AgentSky's suspended machines, but restricted to Claude harnesses on Claude models. AgentSky resells those same Claude models at list price while also offering seven other harnesses and rival models (GPT, Gemini, DeepSeek, Kimi, GLM, Grok), which makes Anthropic both its largest supplier and a single-vendor substitute.
3 products reviewed on TopReviewedModal sells the substrate rather than the agent: sandboxes with gVisor isolation whose compute charges drop to zero while idle, the same suspend-when-idle economics AgentSky bills on. The customer brings their own harness code, model API keys and session plumbing, so Modal is the build-it-yourself alternative to AgentSky's managed harness catalogue.
1 product reviewed on TopReviewedDevin is one proprietary coding agent on its own persistent cloud workspace, reachable from Slack and sold to engineering teams on seats and ACU credits. AgentSky sells the runtime instead of the agent: the caller names the harness and model per request across eight harnesses and eleven models, and pays only for model tokens and awake compute.
1 product reviewed on TopReviewedKey milestones in AgentSky's history, with sources.
First public agent benchmark report covering 40,835 production tasks
AgentSky published "Which agent delivers? The August 2026 rankings", an Elo ranking of eight harness-and-model pairs built from 40,835 real production tasks run between May 6 and August 6, 2026 across $141,676 of measured spend. Claude Code paired with Claude Fable 5 led on quality at Elo 1,201 and a 97.5% completion rate; Codex with GPT-5.6 Sol was named the best value pair at a $1.88 median task cost.
agentsky.devProduct Hunt debut finishes #1 Product of the Day
AgentSky launched on Product Hunt on August 3, 2026 under the tagline "Any harness, any LLM - cloud-hosted agents on demand" and ended the day ranked #1 with 430 upvotes. The launch listing said the platform was already running in production behind tycoon.us, where it had handled more than 10,000 agent sessions.
producthunt.comDeveloper CLI and ACP adapter published to npm
AgentSky published its first public developer tooling on July 31, 2026: @agentsky/sky v0.1.0, the command-line client for launching and talking to long-lived cloud agents from a terminal, followed the same day by @agentsky/acp v0.1.0, an Agent Client Protocol adapter that runs an AgentSky cloud agent as a local ACP agent over stdio.
npmjs.comBrowse multi-perspective AI panel reviews across hundreds of AI tools, agents, and platforms. Find the right software with insights from CTO, Developer, Marketer, Finance, and User perspectives.