Pricing
Transparent usage-based pricing and service tiers for The Swarms API.
Usage-based pricing
Pay only for what you use with transparent, per-operation pricing. All API endpoints use a unified pricing structure for token costs.
Discounts & pricing details
Time-based discounts and unified token pricing across all API endpoints.
Swarm Completions receive a 50% discount on token costs during 8 PM - 6 AM Pacific Time. Agent costs remain the same. Automatically applied.
All API requests are free during Black Friday (4th Friday of November) for 24 hours. Automatically applied.
/v1/usage/costs.Credit deductions
How credits are deducted and billing logic works
API Pricing
Flexible subscription plans that scale with your needs. Choose a plan for higher limits and features; usage is billed per operation. Prices shown are subscription fees; API usage is billed per operation.
Frequently asked questions
Everything you need to know about pricing, billing, and features
How does the usage-based pricing work?
Credits are deducted automatically after each request completes. All API endpoints use unified pricing: $6.50 per 1M input tokens and $18.50 per 1M output tokens. Additional costs apply for agent costs ($0.01 per agent for swarms/workflows), image processing ($0.25 per image), MCP calls ($0.10 per call), Exa Search ($0.04 per search), and Web Scraper ($0.15 per scrape). Free credits are used first, followed by regular credits.
What is unified pricing?
All API endpoints (Swarm Completions, Agent Completions, Advanced Research, Auto Swarm Builder, Graph Workflow, Batched Grid Workflow) use the same token pricing: $6.50 per 1M input tokens and $18.50 per 1M output tokens. This simplifies cost calculation across all endpoints.
How does the night-time discount work?
Swarm Completions receive a 50% discount on token costs during 8 PM - 6 AM Pacific Time. Agent costs remain the same. The discount is automatically applied to eligible API calls during these hours.
What are Premium Endpoints?
Premium Endpoints are advanced features available on Pro and Ultra plans, including batch processing (/v1/agent/batch/completions, /v1/swarm/batch/completions), reasoning agents (/v1/reasoning-agent/completions), batched grid workflows, graph workflows, auto-swarm-builder, and advanced research. These are not available on the Free tier.
What's the difference between Pro and Ultra plans?
Pro ($19.99/month) includes global availability, multi-agent architectures, accelerated hardware, API telemetry, priority support, Pro models, and Premium Endpoints access. Ultra ($100/month) includes everything in Pro plus premium models, higher limits, SOC 2 compliance, enhanced security, priority region/zone routing, and full Premium Endpoints access.
Can I switch between pricing tiers?
Yes! You can upgrade your plan at any time through the Account Dashboard. Go to the Billing tab, scroll to the Plans section, and select your desired plan. You'll be redirected to Stripe to complete payment. Your API keys remain the same, and the upgrade takes effect immediately.
What happens when I run out of credits?
When your credits are exhausted, API calls will fail with appropriate error codes. You can purchase additional credits or upgrade your plan for higher limits. We recommend setting up billing alerts to monitor your usage.
How is image processing charged?
Image processing costs $0.25 per image and is charged when an image is provided to an agent. This applies across all tiers and endpoints.
What is Frenzy Mode?
Frenzy Mode is a special promotion where all API requests are free during Black Friday (4th Friday of November) for 24 hours. The discount is automatically applied - no action needed.
How does the on-premise license work?
The on-premise license ($9,999/year) includes complete Docker deployment, full source code access, unlimited usage, and 1-year enterprise license with priority support. Perfect for organizations requiring data sovereignty and custom deployments.
Start building today
Get started with transparent pricing and pay only for what you use