Solid

Solid is an AI agent platform built on a simple premise: useful agents need more than a chat window. Each Solid agent gets its own computers, its own accounts, and its own budget, so it can take a goal and deliver a finished result rather than a list of suggested next steps. You describe the job in plain language, grant the access it needs, and decide which actions require your approval. The agent works out the steps, connects to the tools involved, builds anything that is missing, checks its own output, and messages you when the work is ready to review or share.

Unusually, Solid does not depend on a catalogue of prebuilt connectors. If a person can use a piece of software, the agent is expected to use it too, whether that means calling an API, writing a missing integration, or operating a website, desktop application, or phone app directly. Agents run across Windows and macOS computers, Linux servers, and iOS and Android devices, and they can create their own Google or Apple accounts when a service demands one. They can also sign up for and pay for services, staying inside the budgets and approval rules you define.

Autonomy is handled in layers the company calls self-healing, self-improving, and self-scaling. When a tool breaks or a setup fails, the agent investigates, repairs what it can, and re-checks that the job still runs; if it stays blocked it explains the obstacle instead of failing silently. Corrections from your team are kept as lessons that shape later jobs, so you should not have to repeat the same instruction twice. For larger jobs, an agent can spin up additional Solid agents or bring in coding agents such as Codex and Claude Code, divide the work, coordinate the group, and return one checked result.

Documented example workflows include turning customer meeting notes into a hosted, tested sales demo; researching LinkedIn leads and drafting outreach for approval before follow-up and CRM updates; building AI product evaluations after each release; resolving a production alert into a reviewed fix; and working a stalled support ticket through to a recorded resolution. Billing runs on monthly credit plans — Mini at $20, Basic at $40, Pro at $160, and Max at $640 — where AI usage, machines, and agent purchases all draw from a single balance with no separate platform fee. A trial is offered, and API access plus enterprise deployment options support teams that want these agents inside their own products, VPCs, or on-premises infrastructure.