Does Trigger.dev support MCP and tool calling?

Full under the Agentic Index verification standard. Trigger.dev publishes public evidence of Model Context Protocol (MCP) support, tool calling, and API or SDK extensibility for connecting agents to external tools. Evidence basis: A TypeScript SDK writing tasks directly in the customer codebase, a CLI, an Apache 2.0 open source core, GitHub and VS Code integration, and an official Model Context Protocol server documented in the developer docs. 2026-08-05. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Does Trigger.dev have SOC 2, SSO and role based access control?

Full under the Agentic Index verification standard. Trigger.dev publishes public evidence of enterprise security and identity controls, covering a compliance posture such as SOC 2 or ISO 27001, single sign on (SSO or SAML), and role based access control (RBAC). Evidence basis: Enterprise plans include a SOC 2 report, role based access control, single sign on and custom log retention, and Bring Your Own Cloud is offered explicitly for teams with strict SOC 2, ISO 27001 or HIPAA requirements. A published security a. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Can Trigger.dev be self hosted or run in a private VPC?

Full under the Agentic Index verification standard. Trigger.dev publishes public evidence of self hosted or private VPC deployment, along with data residency options for controlling where agent data is processed and stored. Evidence basis: The widest deployment range in this pocket: managed cloud, fully self hosted under Apache 2.0 with version tagged releases and a published self hosting guide, or Bring Your Own Cloud running inside the customer's own AWS, GCP or Azure accou. Graded from public evidence only against the Agentic Index 14 point capability taxonomy. No vendor pays for placement.

Back to vendors

Trigger.dev

Q: Who is Trigger.dev for?

TypeScript and backend developers, AI engineering teams, startups and scaleups with compliance obligations

Q: What are the best alternatives to Trigger.dev?

Based on documented capability coverage across the Agentic Index 14 point taxonomy, the closest alternatives to Trigger.dev among agent infrastructure platforms are Inngest, DBOS, Windmill, Agno. Each is graded on the same public evidence, and no vendor pays for placement.

Also known as: Trigger.dev v4, triggerdotdev, Bring Your Own Cloud

Visit site

Agent infrastructureindependentVerified 2026-08-05

Entry priceFree tier reported at around 10,000 runs per month; Pro around $50 per month; Enterprise and Bring Your Own Cloud custom; self hosting free under Apache 2.0Full pricing detail

Open source TypeScript platform for durable AI agents and background workflows with no execution timeouts, checkpointed crash recovery, native human in the loop, an official MCP server, and managed, self hosted or bring your own cloud deployment.

Trigger.dev is an open source platform for building and deploying durable AI agents and background workflows in TypeScript, released under Apache 2.0. Its central promise is the absence of timeouts: tasks run for as long as they need, unlike AWS Lambda, Vercel and other serverless runtimes, and they survive crashes, refreshes and redeploys through checkpointing. Around that sit retries with configurable backoff, queues, idempotency, concurrency control, durable cron style schedules and atomic versioning that lets a new deploy ship without disturbing tasks already running.

For agents specifically it documents the working patterns rather than leaving them to the reader: prompt chaining across multi stage processing, routing work to specialised models based on content analysis, parallelisation of concurrent tasks, orchestrator and worker coordination across multiple agents, and an evaluator and optimiser loop that iteratively refines outputs. Human in the loop is a first class feature of the agent product rather than something bolted on. Realtime connects a frontend directly to a running task so live status and streamed tool calls reach the browser with no polling and no API routes, and it works with the Vercel AI SDK and any model provider.

Two capabilities separate it from the rest of the durable execution pocket. It offers genuine runtime freedom, letting tasks install system packages so a job can run a headless browser, a Python script or FFmpeg, and it can generate code at runtime and execute it inside secure cloud sandboxes. It also publishes an official Model Context Protocol server, documented in its developer docs.

Deployment is the widest in this pocket. Run it on Trigger.dev's managed cloud, self host the Apache 2.0 release on your own infrastructure with version tagged builds, or take Bring Your Own Cloud, which runs the platform inside your own AWS, GCP or Azure account against your existing cloud credits and is aimed explicitly at teams with SOC 2, ISO 27001 or HIPAA obligations. Enterprise plans add a SOC 2 report, role based access control, single sign on and custom log retention. Note that v3 is end of life: 4.5.0 is the last release supporting v3 tasks and later versions reject them.

Vendor details

Canonical URL

https://trigger.dev

Company status

independent

Use cases & customers

Target customers

TypeScript and backend developersAI engineering teamsstartups and scaleups with compliance obligations

Deployment options

managed cloudself-hostedbring your own cloudopen source

In practice

Your agent run legitimately takes forty minutes and every serverless platform kills it. Tasks here have no timeout at all and survive crashes and redeploys through checkpointing.

Your compliance team will not accept a third party running the workload. Bring Your Own Cloud puts the platform inside your own AWS, GCP or Azure account against your existing credits, aimed at SOC 2, ISO 27001 and HIPAA obligations.

The agent needs to drive a headless browser or run FFmpeg mid task. System package support gives real runtime freedom rather than restricting you to a fixed sandbox.

Sources & related URLs

Research sources

https://trigger.dev/ https://trigger.dev/product/ai-agents https://trigger.dev/byo-cloud https://trigger.dev/docs/self-hosting/overview

Agentic Index coverage score

10.5 / 14 capabilities · 75%

Integrations & Tool CallingAny API is reachable through its own SDK or plain fetch, tool calling is supported in agent tasks, and it integrates with GitHub and VS Code and the Vercel AI SDK, but there is no prebuilt connector catalogue or managed authentication layer. 2026-08-05	Partial
Workflow OrchestrationDurable tasks with retries, queues, idempotency, concurrency control, elastic scaling and atomic versioning that deploys without disturbing running tasks, plus documented multi agent patterns covering prompt chaining, model routing, parallelisation and orchestrator worker coordination. 2026-08-05	Full
Knowledge Grounding & RAGNo retrieval, vector store or knowledge grounding capability is provided. Grounding belongs to whichever agent framework or model SDK the developer runs on the platform. 2026-08-05	Unable to verify
Human Oversight & GuardrailsHuman in the loop is documented as a first class capability of the AI agents product rather than an application pattern, alongside an evaluator and optimiser loop that continuously assesses and refines agent output before it is accepted. 2026-08-05	Full
Security, Identity & GovernanceEnterprise plans include a SOC 2 report, role based access control, single sign on and custom log retention, and Bring Your Own Cloud is offered explicitly for teams with strict SOC 2, ISO 27001 or HIPAA requirements. A published security and vulnerability reporting process covers self hosted deployments with version tagged releases. 2026-08-05	Full
Observability & AuditabilityReal time tracing and monitoring of every task run with full observability presented as a core platform pillar, plus log retention and custom dashboards scaling by plan tier and configurable retention at enterprise. 2026-08-05	Full
Memory & State PersistenceTasks are inherently durable through checkpointing, surviving crashes, refreshes and redeploys, and atomic versioning preserves in flight state across deployments so a running agent is not disturbed by a new release. 2026-08-05	Full
Deployment & Data ResidencyThe widest deployment range in this pocket: managed cloud, fully self hosted under Apache 2.0 with version tagged releases and a published self hosting guide, or Bring Your Own Cloud running inside the customer's own AWS, GCP or Azure account against their existing cloud credits. 2026-08-05	Full
Prebuilt Agents, Templates & PacksExample projects and a documented library of agent patterns are published, but there is no prebuilt agent catalogue, template gallery or ready to deploy pack; tasks are written in the customer's own codebase. 2026-08-05	Partial
Triggers & Channel CoverageDurable cron style scheduled tasks without timeout limits, API and batch triggering, an official MCP server as an invocation surface, and Realtime connecting a frontend directly to running tasks for live status without polling. 2026-08-05	Full
Model Flexibility & RoutingWorks with the Vercel AI SDK and any model provider, and documents a routing pattern that distributes tasks to specialised models based on content analysis, but that is an implementation pattern rather than a platform level model registry, routing policy or key management surface. 2026-08-05	Partial
APIs, SDKs & MCP ExtensibilityA TypeScript SDK writing tasks directly in the customer codebase, a CLI, an Apache 2.0 open source core, GitHub and VS Code integration, and an official Model Context Protocol server documented in the developer docs. 2026-08-05	Full
Testing, Debugging & OptimizationReal time tracing supports debugging and a documented evaluator and optimiser pattern iteratively refines agent output, plus deploy rollback from the dashboard, but no evaluation harness, benchmark suite or replay testing framework is documented. 2026-08-05	Partial
Browser & Computer UseTasks can install system packages and run headless browsers, Python scripts or FFmpeg, and the platform executes runtime generated code inside secure cloud sandboxes, so it HOSTS computer use workloads. It provides no browser control primitive of its own, which is why this is partial rather than full. Distinguishes it from inngest and pipedream, both graded N. 2026-08-05	Partial

The Agentic Index coverage score grades every vendor Full, Partial or Unable to verify against the same 14 buyer facing capabilities, from public evidence only. Each capability links to how all vendors in the index score on it. How this evidence is graded

Pricing

Free tier reported at around 10,000 runs per month; Pro around $50 per month; Enterprise and Bring Your Own Cloud custom; self hosting free under Apache 2.0

runs and compute usage

Free tierTrial available

What is public

Plan structure, the Apache 2.0 self hosting route and the enterprise feature list are public; specific rates circulate mainly through third party comparisons.

Billing mechanics

Managed cloud tiers metered on runs and compute, with a free plan for small projects and a Pro tier for production use. Enterprise is custom and carries the governance features. Separately, the Apache 2.0 licence permits unlimited free self hosting where the only cost is the customer's own infrastructure, and Bring Your Own Cloud is a managed experience running inside the customer's cloud account using their existing AWS, GCP or Azure credits.

Cost watchouts

The SOC 2 report, role based access control, single sign on and custom log retention are all enterprise tier, so a compliance requirement moves a buyer off the published rates entirely

Variable cost rationale

Managed tiers meter runs and compute, and no timeout limits mean a single agent task can consume far more compute than a background job, but the Apache 2.0 self hosting and Bring Your Own Cloud routes both give a way to cap vendor spend by absorbing execution into infrastructure the team already runs.

Additional watchouts

Compliance features are gated to enterprise, so SOC 2, RBAC and SSO cannot be bought self serve. Version management also matters commercially: v3 is end of life, 4.5.0 is the last release supporting v3 tasks, and later versions reject v3 triggers and deploys, so a migration may be a prerequisite for anyone on an older deployment.

Overage / add-ons

Usage above plan allowances is metered on runs and compute; higher volume or compliance needs move the customer to enterprise terms

Sales call required

Mixed (some tiers require a call)

Free / trial

Free plan reported at roughly 10,000 runs per month, plus a fully free self hosted route under the Apache 2.0 licence with no usage limits imposed by the vendor

Lowest paid plan

Pro

Commercial notes

The Bring Your Own Cloud route is commercially distinctive in this pocket. Rather than choosing between a managed service and operating the platform yourself, a compliance bound team gets vendor operated software running on infrastructure it already pays for, which also lets committed cloud spend absorb the compute. Temporal offers self hosting and Inngest keeps compute on customer endpoints, but neither packages this middle option the same way.

Key ambiguities

The free tier run allowance and the Pro rate come from third party comparisons rather than a retrieved first party rate card, and Bring Your Own Cloud pricing is explicitly quote only. Re verify at trigger.dev/pricing before publishing figures.

Missing data

Confirmed first party rates, the exact free tier allowance, enterprise pricing, and Bring Your Own Cloud pricing.

Agentic Index verified 2026-08-05

Official pricing page

Related vendors

Alternatives to Trigger.dev

The closest documented capability profiles to Trigger.dev among agent infrastructure platforms tracked by Agentic Index, ordered by similarity on the same 14 point evidence the rankings use. No vendor pays for placement.

Inngest10.5 / 14Fuller documented coverage on Testing, Debugging & Optimization
DBOS9.0 / 14A lighter documented profile than Trigger.dev
Windmill10.0 / 14Fuller documented coverage on Integrations & Tool Calling
Agno12.5 / 14Adds documented Knowledge Grounding & RAG
Temporal9.5 / 14Fuller documented coverage on Integrations & Tool Calling
Tray.ai11.5 / 14Adds documented Knowledge Grounding & RAG

Similarity is computed from each vendor's Agentic Index coverage score evidence, axis by axis, not from the totals. How this evidence is graded

Compare Trigger.dev with other vendors →Compare all agent infrastructure platforms →See where Trigger.dev ranks →

Data confidence: high