@scroll.window="scrolled = (window.scrollY > 10)">

Pricing

The Right Plan for Your Team

Start building for free, then scale as you grow. No credit card required to get started.

Sandbox (Free)

For developers and architects to prove the value of deterministic context engineering.

$0 /forever

Start for Free
  • Unlimited Agents & Team Members
  • Included 100k AI Credits / month
  • Inspector Console included (Unlimited audit logs)
  • Teaser access to Deep Context (Graph) & Analytics (~50 runs/day)
  • Ingestion rate limits (up to ~50 files/day)
  • Best-effort performance on shared ingestion queue
  • Audio/Video Ingestion
  • Dedicated & Isolated Indexes
  • Community Support

Production

Enterprise Grade

For regulated deployments requiring audit logs, extended retention, and high throughput.

$149 /month

Upgrade to Production
  • Everything in Free, plus:
  • 15 Million AI Credits / month Enough for heavy production workloads
  • 90-Day Audit Log Retention
  • High-throughput ingestion capacities
  • Priority, GPU-accelerated ingestion queue for faster processing
  • Full access to Context Engine (Graph) & Aggregate
  • Full audio & image/OCR ingestion
  • High-precision reranker
  • Dedicated & isolated indexes
  • Priority support

Enterprise

For large-scale deployments with custom security, support, and deployment needs.

Custom

Contact Sales
  • Everything in Pro, plus:
  • Unlimited* AI Credits
  • Custom Credit Allocation
  • On-Premise Deployment Option
  • White-Label & Custom Branding
  • Dedicated Account Manager
  • Dedicated Support & SLA

* “Unlimited” credits available only under Enterprise agreements with Fair Use policy, contractual rate limits/concurrency, and SLA. Contact sales for details.

Frequently Asked Questions

An AI Credit is the unit of measure for your AI agent's activity on DhakaBrain. Credits are only consumed when an agent (or a backend service) calls our core reasoning engines: /mcp/context (Smart Context) and /mcp/aggregate (Analytics). We meter the work done (complexity of the graph or aggregation), not just storage.

The Context Engine smartly routes requests: simple lookups cost less (Retrieve price), while complex reasoning chains cost more (Graph price). You only pay for the intelligence you use.

No surprises. If you go over your plan's included usage, requests may start to return rate-limited (429) responses instead of silently failing or auto-charging you. You can then either upgrade to the Pro plan, wait for the monthly counters to reset or purchase additional AI Credits from the in-app billing page. We'll always surface clear messages when you're approaching or hitting a limit.

Yes. We believe developers should be able to test and debug without limits. That's why all manual usage of the Cognitive Console by a human does not consume your AI Credits on any plan.
On the Pro plan, this means all console use is truly unlimited. On the Free plan, standard retrieval is unlimited, while advanced cognitive tools (Aggregate & Graph building) are exposed with generous "teaser" limits (~50 runs/day) to let you test the power of the engines before upgrading.

The Free plan is designed for you to explore, learn, and build your first proof-of-concept agents with unlimited team members, console use, and fair-use AI Credits, within tighter ingestion limits (files and volume per day). The Pro plan is built for production applications, with significantly higher usage allowances, access to all Pro tools (Aggregate, TaskGraph, Reranker), higher-throughput ingestion (including audio/image/OCR), and a more isolated infrastructure path suitable for security-sensitive workloads.

Yes. Our Partner Program is designed for agencies and consultants who build solutions for multiple clients. It includes features like multi-tenant organization management, white-labeling options, and access to AI Credits at a significant discount. It's a custom plan, so please contact our sales team to discuss a partnership.

How far do 15M credits go?

A practical rule-of-thumb for sizing production agents.

In practice, 15M monthly credits give you room for several agents running steady production traffic without constantly watching the meter.

  • Rough rule-of-thumb: about 1M credits / month is typically enough for a single production agent with regular workloads.
  • 5M credits can usually cover 10–15 agents, depending on how often they call /mcp/context and /mcp/aggregate.
  • Spiky workloads or very large contexts will consume more credits; lighter agents will consume less.

These numbers are indicative only and may vary by use case. You can always add extra credits or move to a custom plan if you outgrow the included 15M.