CrewAI — Free for 50 executions a month, then enterprise — the $25 middle tier was removed in spring 2026, leaving nothing between $0 and a five-figure contract.
Short answer: CrewAI orchestrates multi-agent workflows. As of 2026 it offers two tiers: Basic is free with 50 workflow executions a month, one seat, a visual editor and AI copilot; Enterprise is custom-quoted with executions sized to workflow and flexible overage, reported at $60,000 to $120,000 a year with up to 30,000 executions. A $25 Professional tier existed from October 2025 until it was removed in spring 2026. Billing is per execution, not per token or seat.
On this page
Verdict at a glance
Overall rating: 3.5/5 ★★★☆☆
Standout: Execution-based billing. One execution is one complete crew kickoff regardless of internal complexity, token consumption or duration — a genuinely simple unit.
Watch out: The $25 Professional tier was removed in spring 2026, so there is now nothing between a free 50-execution plan and an enterprise contract reported at $60,000 a year.
CrewAI orchestrates multiple AI agents working together on a task, and its billing unit is unusually clean: one execution equals one complete crew kickoff, regardless of internal complexity, token consumption or duration. That means a workflow involving six agents and forty tool calls costs the same as one involving two agents and three — you are billed for the job, not the machinery. Compared with token metering, which punishes complexity you cannot easily predict, it is a considerably more forecastable model. The commercial structure is the problem. As of 2026 there are two tiers: Basic is free with 50 workflow executions a month, one seat, a visual editor and an AI copilot; Enterprise is custom-quoted with executions sized to workflow and flexible overage, reported at $60,000 to $120,000 annually for organisations needing compliance certifications and dedicated support, including up to 30,000 executions. A $25 a month Professional tier providing 100 executions with $0.50 per additional execution existed from the October 2025 launch of AMP until it was removed in spring 2026. Its removal leaves nothing at all between free and a reported five-figure contract, which is a very large gap for any team that outgrows 50 monthly executions but is nowhere near enterprise. Note also that LLM API costs are yours, not CrewAI’s.
Quick facts
| Attribute | Detail |
|---|---|
| Ideal team | Teams orchestrating multi-agent workflows who fit free or enterprise |
| Learning curve | Moderate — a visual editor with real underlying complexity |
| Setup time | Hours for a first crew |
| Free trial | Free Basic tier |
| Free tier | Yes — 50 workflow executions monthly, one seat |
| Billing unit | ⭐ Executions — one complete crew kickoff |
| Removed tier | ⚠️ $25 Professional existed Oct 2025 to spring 2026 |
| Key limit | ⚠️ Nothing between free and reported $60,000+ enterprise |
Overview
CrewAI coordinates multiple AI agents working together on a task — assigning roles, passing context between them and sequencing their work toward an outcome. Its billing unit is the clearest in this comparison: one execution equals one complete crew kickoff, regardless of internal complexity, token consumption or duration. A workflow with six agents and forty tool calls costs one execution, exactly like a workflow with two agents and three calls. Against token metering, which makes cost a function of complexity you cannot easily estimate in advance, that is far more forecastable.
The tier structure as of 2026 has two levels. Basic is free and provides 50 workflow executions a month with one seat, a visual editor and an AI copilot — enough to build and test crews properly. Enterprise is custom-quoted with executions sized to the workflow and flexible overage negotiated per contract, reported at $60,000 to $120,000 annually for organisations requiring compliance certifications and dedicated support, and including up to 30,000 executions.
What sits between them is nothing, and that is a recent change worth knowing. A Professional tier at $25 a month providing 100 executions, with $0.50 per additional execution, existed from the October 2025 launch of AMP until it was removed in spring 2026. Its withdrawal leaves a team that outgrows 50 monthly executions with no purchasable option short of an enterprise conversation — a gap of roughly three orders of magnitude. That is a structural problem for small and mid-sized teams rather than a pricing quibble. The other cost to account for is that CrewAI bills executions but does not cover the LLM API calls those executions make; model spend is yours and is frequently the larger number.
Key features
- Multi-agent orchestration — roles, context passing and sequencing
- Execution-based billing — ⭐ one crew kickoff regardless of complexity
- Visual editor — included on the free Basic tier
- AI copilot — included on Basic
- 50 free executions monthly — Basic tier
- Up to 30,000 executions — reported on Enterprise
- Flexible overage — negotiated per Enterprise contract
- Compliance certifications — Enterprise tier
- Dedicated support — Enterprise tier
- Self-hosting — available for the open-source framework
Pricing (as of 2026-09-28)
Model: execution-metered, with only a free tier and a custom enterprise tier · Free tier: yes — 50 workflow executions monthly, one seat · Entry paid: ⚠️ Enterprise only — reported $60,000-$120,000 a year
| Plan | Price | Includes |
|---|---|---|
| Basic | ⭐ $0 | 50 workflow executions per month · 1 seat · visual editor · AI copilot |
| Enterprise | ⚠️ Custom — reported $60,000-$120,000 a year | Executions sized to workflow with flexible overage negotiated per contract · reported to include up to 30,000 executions · compliance certifications · dedicated support |
| Professional (removed) | ⚠️ Was $25/month | 100 executions with $0.50 per additional execution · existed from the October 2025 AMP launch until removed in spring 2026 |
| LLM API costs | ⚠️ Not included | CrewAI bills executions but does not cover the model API calls those executions make — model spend is yours and is often the larger figure |
⚠️ The $25 Professional tier providing 100 executions was removed in spring 2026, leaving nothing between a free 50-execution plan and an enterprise contract reported at $60,000 to $120,000 a year. For any team that outgrows the free tier but is nowhere near enterprise, there is currently no purchasable option.
⭐ Billing is per execution rather than per token, API call or seat — one execution is one complete crew kickoff regardless of internal complexity, token consumption or duration. That makes cost far more forecastable than token metering, where complexity you cannot predict drives the bill.
⚠️ CrewAI does not cover the LLM API calls your executions make. Model spend is billed separately by your provider and is frequently the larger number, so the execution allowance describes only part of what running crews actually costs.
Ratings breakdown
| Criterion | Score | Why |
|---|---|---|
| Ease Of Use | 4.0/5 | Visual editor and copilot make crew building approachable |
| Features | 4.0/5 | Solid multi-agent orchestration with a clean billing unit |
| Value | 3.0/5 | ⚠️ Excellent free tier, then a three-order-of-magnitude gap |
| Support | 3.5/5 | Community on Basic; dedicated support at Enterprise |
Pros
- Execution billing is the cleanest unit in this comparison
- One execution covers a whole crew kickoff regardless of complexity
- Free Basic tier provides 50 monthly executions with a visual editor
- AI copilot included on the free tier
- Cost is forecastable in a way token metering is not
- Enterprise reported to include up to 30,000 executions
- Compliance certifications and dedicated support at Enterprise
- Open-source framework available for self-hosting
Cons
- ⚠️ The $25 Professional tier was removed in spring 2026
- ⚠️ Nothing purchasable between free and reported $60,000 enterprise
- ⚠️ LLM API costs are separate and often larger than execution fees
- Basic limited to one seat
- Enterprise pricing is custom with no published reference
- Fifty free executions is quickly exhausted by real testing
- Teams outgrowing the free tier have no upgrade path short of sales
Who it’s for
- Teams whose usage genuinely fits within 50 monthly executions
- Enterprises needing compliance certifications and dedicated support
- Developers who prefer execution billing to token metering
- Organisations self-hosting the open-source framework
Who it’s not for
- Small and mid-sized teams outgrowing 50 executions — no tier exists
- Anyone budgeting from execution costs alone, given separate LLM spend
- Teams needing more than one seat on a free plan
- Organisations wanting published pricing before procurement
Best use cases
- Orchestrating several agents on a task with roles and context passing
- Building and testing crews within 50 free monthly executions
- Forecasting cost per workflow rather than per token consumed
- Running up to 30,000 executions on a reported enterprise contract
- Self-hosting the open-source framework outside the managed platform
Integrations
CrewAI connects to the major model providers and tool ecosystems, with the important caveat that it orchestrates those calls rather than paying for them — your LLM provider bills you directly for the inference each execution consumes. The open-source framework can be self-hosted, which is the route for teams who need CrewAI’s orchestration without the managed platform’s pricing gap.
How it compares
| CrewAI | LangChain | Beam AI | Decagon | |
|---|---|---|---|---|
| Entry paid | ⚠️ Enterprise only | $39/seat/mo (LangSmith Plus) | $50/mo (Pro) | ⚠️ $50,000/yr platform fee |
| Free tier | ⭐ 50 executions monthly | ⭐ Framework free, 5,000 traces | ⭐ Yes — 20 tasks a month | ⚠️ None |
| Mid-tier available | ⚠️ No — removed spring 2026 | ⭐ Yes at $39 | Pro $50/mo, then Scale $3,990 | No |
| Billing unit | ⭐ Executions | Traces | Tasks per month | Conversations |
| LLM costs included | ⚠️ No | ⚠️ No | Bundled | Bundled |
| Self-hostable | ⭐ Open-source framework | ⭐ Framework open source | Enterprise VPC | No |
| Best fit | Multi-agent orchestration | LLM app tracing | Agentic process automation | Support conversations |
LangSmith figures are from published third-party reporting verified 2026-09-29. Sierra, Decagon and Beam AI publish no public pricing, and CrewAI publishes only its free Basic tier.
Alternatives
LangChain is the code-first alternative with a free open-source framework and a $39 per seat observability layer — and it has the mid-tier CrewAI removed. Beam AI handles agentic process automation at reported tiers from $990. Sierra and Decagon are customer-service specific rather than general orchestration.
Recent changes
As of 2026 CrewAI offers two tiers. Basic is free with 50 workflow executions per month, one seat, a visual editor and an AI copilot. Enterprise is custom-quoted with executions sized to workflow plus flexible overage negotiated per contract, estimated at $60,000 to $120,000 annually for organisations requiring compliance certifications and dedicated support, and reported to include up to 30,000 executions. A Professional tier priced at $25 a month with 100 executions and $0.50 per additional execution existed from the October 2025 launch of AMP until it was removed in spring 2026, leaving no purchasable option between the free and enterprise tiers. CrewAI bills based on executions rather than API calls, tokens or seats, with one execution equal to one complete crew kickoff regardless of internal complexity, token consumption or duration. LLM API costs are not covered by CrewAI and are billed separately by the model provider.
Frequently asked questions
How much does CrewAI cost?
Basic is free with 50 workflow executions a month and one seat. Enterprise is custom-quoted and reported at $60,000 to $120,000 a year with up to 30,000 executions. There is currently nothing in between — the $25 Professional tier was removed in spring 2026.
What happened to CrewAI’s $25 plan?
The Professional tier, which provided 100 executions a month with $0.50 per additional execution, existed from the October 2025 AMP launch until it was removed in spring 2026. Its withdrawal leaves nothing purchasable between a free 50-execution plan and a reported five-figure enterprise contract.
What counts as a CrewAI execution?
One complete crew kickoff, regardless of internal complexity, token consumption or duration. A workflow with six agents and forty tool calls is one execution, exactly like one with two agents and three calls. That makes cost far more forecastable than token metering, where unpredictable complexity drives the bill.
Does CrewAI cover my LLM costs?
No. CrewAI bills executions but your model provider bills you separately for the inference those executions consume — and that is frequently the larger number. The execution allowance therefore describes only part of what running crews actually costs, so model both before comparing against a platform that bundles inference.
What if I outgrow CrewAI’s free tier?
Currently there is no answer short of an enterprise conversation, since the mid-tier was removed. That is a genuine structural problem for small and mid-sized teams. LangChain offers a free framework with a $39 per seat observability layer, which is the closest thing to the tier CrewAI withdrew.
CrewAI or LangChain?
CrewAI for visual multi-agent orchestration with clean execution billing, if your usage fits the free tier or you are buying enterprise. LangChain for code-first building with a free open-source framework, the largest ecosystem, and a mid-tier that actually exists at $39 per seat.
Can I self-host CrewAI?
The open-source framework can be self-hosted, which is the practical route for teams caught in the pricing gap — you get the orchestration without the managed platform’s tiers. You still pay your own LLM provider for inference, and you take on the operational work the managed service would otherwise handle.
Sources
- CrewAI — pricing (vendor)
- Lindy — CrewAI pricing, features and alternatives explained 2026
- ZenML — CrewAI pricing guide: plans and features the framework offers
- Coverge — LangSmith pricing in 2026: tiers, costs and what to watch for
- AIChief — Beam AI review: cost, use cases and alternatives 2026
- eesel AI — Decagon AI pricing in 2026: what it actually costs
Compiled from public sources; last verified 2026-09-28. Pricing is subject to change — verify with the vendor. Ratings reflect editorial opinion based on documented capabilities, published pricing, and aggregated user reviews. DigiJaws does not accept payment for review placement or rating.
CrewAI Review (3.5/5): Pricing & Verdict (2026). DigiJaws, 2026. https://digijaws.com/reviews/crewai-review/