Metorial vs Toolhouse (2026)

More than 1,000 catalog integrations are available, each deployed as a provider with its own config and per user authentication (OAuth through Provider Setup Sessions, API tokens or custom credentials). Their tools reach the agent through a session's MCP URL or direct tool calls on the API. Customers add their own tools as custom providers deployed from code, container images or remote MCP servers, and can use their own OAuth apps.

Sourcemetorial.com/docs/build/operate/providersread 2026-09-22

Workers connect to more than a thousand services through built in integrations Toolhouse maintains and through integrations that need a connection, authorized per service over OAuth through Auth Connections, with the actions a worker may perform chosen per integration. Workers also call tools on remote MCP servers over SSE or streamable HTTP with OAuth, and can call third party APIs from uploaded API documentation in the virtual computer.

SourceToolhouse, docs.toolhouse.ai toolhouse/integrations, auth-connections and mcp-serversread 2026-09-21

Agents connect to tools through Metorial, which does not run the agent itself. Sessions merge the tools of several providers behind one connection. Magic Skills package instructions, files and optional scripts that the customer's AI client or agent follows when it invokes the skill. Sequencing and branching happen in that client or in the customer's framework, and Metorial has no workflow runtime of its own.

Sourcemetorial.com/docs/platform/workforce/magic-skillsread 2026-09-22

A manager worker coordinates specialized workers through the Call Agent integration, deciding which to involve, running them in sequence, in parallel or mixed according to data dependencies, passing one worker's output to the next, and chaining managers to any depth. Workers can also trigger other workers as a pipeline. The customer configures that routing across steps.

SourceToolhouse, docs.toolhouse.ai capabilites/worker-orchestrationread 2026-09-21

Providers expose triggers that fire on events in the external system. A callback, which is a webhook Metorial registers against every matching integration instance, notifies the customer's application so it can run event driven workflows. Callbacks are a metered plan feature, with 10,000 a month on Dev and 100,000 on Scale.

Sourcemetorial.com/docs/build/glossaryread 2026-09-22

Schedules start a worker autonomously on a cadence set in natural language or in the interface, as often as every ten minutes and with several schedules per worker, and workers also start on events such as a webhook notification from another service, so work reaches the agent without a person initiating it. Each worker also has its own email inbox, a chat interface and an API endpoint, with an optional completion webhook.

SourceToolhouse, docs.toolhouse.ai toolhouse/schedules and introduction-to-ai-workersread 2026-09-21

Agents get live access, through Metorial's integrations, to the systems that hold company knowledge, such as Google Drive, Notion and Confluence. A Magic Skill can carry reference documents for its instructions. There is no index, retrieval layer or semantic search over the customer's corpus. The API has Documents and Stores resources, but Metorial does not present them as retrieval.

Sourcemetorial.com/docs/platform/get-started/core-conceptsread 2026-09-22

Agent Files attach knowledge to a worker, such as guidelines, templates, skill files and reference documents in text or PDF, and their contents are added directly to the worker's instructions, where they persist for every run. The document parser and file download bring further files in for each task. There is no index, graph or embeddings layer over the customer's knowledge that scales past the context window and stays queryable.

SourceToolhouse, docs.toolhouse.ai capabilites/skills-and-knowledge and capabilites/document-parserread 2026-09-21

Sessions are lightweight runtime contexts created per request or per user interaction. Metorial stores each user's provider credentials as auth configs for reuse. There is no memory layer that keeps an agent's context across runs.

Sourcemetorial.com/docs/build/operate/sessionsread 2026-09-22

Workers write memories of data, behavior and past outputs through a Memory Store integration and read the relevant ones back through Semantic Memory Search on later runs, with reading and writing separable so a worker can be read-only, and each run keeps its full conversation history under a run ID across chat, email and the API. Toolhouse states no lifetime for stored memories and offers no surface to review, edit or delete them.

SourceToolhouse, docs.toolhouse.ai capabilites/memory and developers/worker-statefulnessread 2026-09-21

Several controls constrain what agents may do. Access groups decide which integrations and skills each user or SSO group sees. People request more access, and administrators review and resolve those requests. Session tool filters can restrict a provider to read only tools, and Protoguard screens each message and tool call, blocking prompt injection and requests outside the customer's policies. No person approves a specific agent action before it commits.

Sourcemetorial.com/docs/platform/workforce/portalsread 2026-09-22

Agent Builder writes a governance section into each worker's instructions stating what controls apply and when it should stop and escalate, for example drafting a campaign and requiring approval before sending, and a worker can start autonomously and email its owner for approval before continuing. The approval lives in the worker's prompt, not in a gate the platform enforces before an action runs, so the boundary is stated, not enforced.

SourceToolhouse, docs.toolhouse.ai introduction-to-ai-workers and toolhouse/worker-emailsread 2026-09-21

The platform is SOC 2 Type 2 compliant, independently audited for security, availability and confidentiality, with reports under NDA. It is also GDPR compliant, with a DPA and a published subprocessor list, and Metorial displays a SOC 2 badge on its homepage. Access uses SSO and SAML with per user and per group policies, and roles are named permission sets attached to members and teams. SCIM and SAML IdP sign in are offered too, and secrets are encrypted in Vault with keys managed through KMS. RBAC and SAML SSO come with the Enterprise plan.

Sourcemetorial.com/securityread 2026-09-22

Workers reach only the services and actions they are given, through OAuth auth connections scoped per worker and action selection per integration, a least privilege model for tool access. A worker can be kept private to its owner, API keys are encrypted in transit and at rest, customer data is encrypted, not used for model training and scoped to the workspace, and Enterprise customers can arrange custom authentication models. There is no identity provider integration such as SSO or SAML and no attestation or trust center, and the toolhouse.ai footer links only privacy and terms pages.

SourceToolhouse, docs.toolhouse.ai auth-connections, developers/authentication and toolhouse.ai/pricingread 2026-09-21

Connection logs record what agents did. Each session has a timeline, and every tool call opens to its tool name, arguments and result. Logs filter across sessions by status, source and agent, with tool errors listed separately. Workforce shows each agent's operations and connection history tied to its linked identity. The logs also carry provider auth and Protoguard events, and the Tool Calls API exposes inputs, outputs and status. Log retention is 30 days on Dev, 90 on Scale and custom on Enterprise.

Sourcemetorial.com/docs/platform/integrations/review-connection-logsread 2026-09-22

The Logs page records every worker execution with status, run and user IDs and duration, a trace of each integration call in order with a timing waterfall and its exact input and output, and a history of the worker's intermediate outputs and any approvals or directions a person gave, with log retention stated by plan: 30 days on Business, one year on Business Max and custom on Enterprise.

SourceToolhouse, docs.toolhouse.ai toolhouse/logs and toolhouse.ai/pricingread 2026-09-21

Deployment can sit in Metorial's cloud or the customer's. On premises deployment runs in the customer's own cloud or data center, and customers can choose EU or US regions for data residency. Each provider runs in its own isolated enclave behind firewalls. On premises deployment comes with the Enterprise plan.

Sourcemetorial.com/securityread 2026-09-22

Toolhouse is a hosted service on every plan, Business, Business Max and Enterprise, with no self hosted, private cloud, on premises or region selection option.

SourceToolhouse, toolhouse.ai/pricing and docs.toolhouse.ai; toolhouse.ai/en/pricingread 2026-09-21

Metorial supplies the tools to package a customer's own workflows. Magic Skills come with templates, groups and curated skill marketplaces, imports from public repositories, and portals that publish them to users. The skills in those catalogs are written by the customer's own experts. The integration catalog is a set of integrations, and the three sample apps in the docs are tutorial code. Metorial ships no ready made skill, template or role specific agent of its own.

Sourcemetorial.com/docs/platform/workforce/magic-skillsread 2026-09-22

Agent Builder assembles each worker from built in capabilities Toolhouse maintains, among them scraping, web and social search, memory, image generation, vision, a virtual computer and document parsing, and common worker patterns such as briefing, triage and content creation are ones the customer builds from a prompt. Workers an organization has built can be shared with its team. Toolhouse offers no catalog or named set of ready made workers a customer adopts, so these are assets the customer assembles.

SourceToolhouse, docs.toolhouse.ai readme and introduction-to-ai-workersread 2026-09-21

Tools reach whatever model the customer runs. The provider runtime formats session tools for OpenAI, Anthropic, Google Gemini and OpenAI compatible APIs such as DeepSeek, Together AI and xAI, and for the Vercel AI SDK, each called with the customer's own provider key. Any MCP client, such as Cursor, Claude Code or Codex, can connect. Metorial names no model of its own, so the model, provider and key are always the customer's choice.

Sourcemetorial.com/docs/build/providersread 2026-09-22

Every plan includes model usage, which Toolhouse calls free unlimited tokens. The models or providers workers run on are not named, the customer cannot choose one, and nothing describes working across models, routing between them, or bringing one's own provider and keys.

SourceToolhouse, toolhouse.ai/pricing and docs.toolhouse.ai; toolhouse.ai/en/pricingread 2026-09-21

A REST API with keys scoped to one environment makes Metorial callable from outside. It covers providers, deployments, sessions, tool calls, skills, portals and callbacks, alongside JavaScript and Python SDKs and a CLI. Metorial serves MCP as addressed endpoints, a session connection URL and Magic MCP URLs. Customers package their own tools as versioned custom providers with rollback.

Sourcemetorial.com/docs/build/apiread 2026-09-22

Every worker is callable as an API at its own endpoint, authenticated with a Toolhouse API key. POST starts a task, PUT continues it under a run ID, GET returns the full history, NDJSON endpoints stream tool calls alongside the output, and long virtual computer jobs run in the background for polling.

SourceToolhouse, docs.toolhouse.ai developers/workers-api and developers/authenticationread 2026-09-21

Customers can test before production. Sandbox environments have their own deployments, keys, auth configs and logs, and a development key never reaches production. Metorial Explorer calls a provider's tools against test data, and custom providers have version pinning and instant rollback. There is no way to evaluate the customer's agent with a readable, comparable result such as scored test cases, a pass rate or a judge verdict.

Sourcemetorial.com/docs/build/operate/organizations-and-projectsread 2026-09-22

The Testing Workbench runs a new or edited worker on real test inputs so its builder can check the output before use and correct the worker through Agent Builder. There are no scored test cases, datasets, evaluation harness or quality gate that returns a verdict.

SourceToolhouse, docs.toolhouse.ai toolhouse/quickstart and introduction-to-ai-workersread 2026-09-21

Everything runs through APIs and MCP servers, which is how Metorial's providers reach applications. Metorial has no browser, desktop or remote machine control of its own.

Sourcemetorial.com/docs/build/operate/providersread 2026-09-22

A built in scraper that Toolhouse maintains, attached when a task needs it, reads web pages at run time with location selection by country, and Vision can capture and read a page screenshot. Toolhouse says the scraper reads pages and does not fill forms or log in, which is web scraping, not an agent operating a browser. The virtual computer executes Python code and operates no interface.

SourceToolhouse, docs.toolhouse.ai capabilites/scraping and capabilites/visionread 2026-09-21