Free
$0/ month
5,000 decisions
Start building with Owctopus.
- 5,000 decision executions/month
- SDK access
- API access
- Basic conditions
- AI fallback
- Basic decision logs
- Development environment
- Basic dashboard
Bring your own AI key
Start freePricing
Pay for decisions. Measure the savings.
You pay Owctopus for decision infrastructure. You still pay your AI provider for unknown decisions.
The question that matters is net savings: avoided model calls minus Owctopus cost. Bring your own AI key. Start free, then scale by volume. Shown in USD by default; INR for India.
Sized by volume.
Every plan includes the SDK and your own AI key. Validation stays available — learning tools scale with the plan.
Developer
$12/ month
50,000 decisions
For developers putting decision intelligence into production.
- 50,000 decision executions/month
- Production API
- API keys
- Decision history
- Conditions
- AI fallback
- Basic analytics
- Decision explanations
- Outcome tracking
- Multiple API keys
Most popular
Growth
$59/ month
500,000 decisions
For startups building decision-heavy products.
- 500,000 decision executions/month
- Everything in Developer
- Pattern discovery
- Candidate conditions
- Condition validation
- Outcome learning
- Decision analytics
- AI cost analytics
- Simulation
- Decision versioning
- Multiple environments
- Learning insights
- Decision maturity
Scale
$239/ month
3,000,000 decisions
For high-volume production decision systems.
- 3,000,000 decision executions/month
- Everything in Growth
- Advanced learning
- Advanced analytics
- Advanced decision traces
- Higher rate limits
- Webhooks
- Team access
- Advanced retention
- Priority support
- Advanced simulation
- Advanced condition management
- SLA options
Enterprise
Custom
10M+ decisions
For mission-critical decision infrastructure.
- 10M+ decisions
- Dedicated infrastructure
- Private networking
- Enterprise security
- SSO
- Audit logs
- Custom data retention
- Custom SLA
- Dedicated support
- Custom AI routing
- Dedicated environments
- Data residency options
- Custom rate limits
Bring your own AI key
Talk to usEstimate your Owctopus ROI
Pick a published OpenAI or Claude model rate, your typical tokens per call, and condition hit rate — then see whether decision reuse pays for itself.
500,000
5k1.5M3M
60%
0%50%100%
300,000 condition hits / month under this assumption
Choose a provider model. Rates are published USD list prices per 1M tokens (as of 2026-10-09).
Estimated cost per avoided call: $0.0004 ($0.0004 at provider rates)
Plan cost uses the published monthly price plus any overage for this volume.
Optional allowance for added infrastructure or operations.
Estimated monthly outcome
Under these assumptions, estimated avoided AI costs exceed the listed Owctopus and additional costs.
- Model
- OpenAI GPT-4o mini
- Tokens / call
- 1,200 in · 300 out
- Recommended / selected plan
- Growth
- Owctopus plan + overage
- $59
- Other costs
- $0
- Condition hit rate
- 60%
Illustrative calculator, not a prediction or measured benchmark. Model rates are the providers' published standard API prices (OpenAI / Anthropic), not a live billing API. Assumes every condition hit avoids one otherwise necessary AI call. Does not include prompt-cache discounts, batch pricing, learning/validation AI, Free-tier overage, or plan-specific limits.
Bring your own AI key
Your model bill stays with your provider.
Connect your own OpenAI-compatible key. Owctopus uses it only when a decision is still unknown. The key is encrypted at rest.
What each plan includes.
| Feature | Free | Developer | Growth | Scale | Enterprise |
|---|---|---|---|---|---|
| Decision executions | 5,000 | 50,000 | 500,000 | 3,000,000 | 10M+ |
| SDK | ✓ | ✓ | ✓ | ✓ | ✓ |
| API | Development | Production | ✓ | ✓ | ✓ |
| Conditions | Basic | ✓ | ✓ | Advanced | ✓ |
| AI fallback | ✓ | ✓ | ✓ | ✓ | ✓ |
| Decision history | Basic logs | ✓ | ✓ | Advanced traces | ✓ |
| Analytics | — | Basic | ✓ | Advanced | ✓ |
| Outcome tracking | — | ✓ | ✓ | ✓ | ✓ |
| Pattern discovery | — | — | ✓ | ✓ | ✓ |
| Condition learning | — | — | ✓ | Advanced | ✓ |
| Simulation | — | — | ✓ | Advanced | ✓ |
| Decision versioning | — | — | ✓ | ✓ | ✓ |
| Multiple environments | Development | — | ✓ | ✓ | Dedicated |
| AI cost analytics | — | — | ✓ | ✓ | ✓ |
| Decision maturity | — | — | ✓ | ✓ | ✓ |
| Advanced traces | — | — | — | ✓ | ✓ |
| Webhooks | — | — | — | ✓ | ✓ |
| Teams | — | — | — | ✓ | ✓ |
| Priority support | — | — | — | ✓ | Dedicated |
| SLA | — | — | — | Options | Custom |
| Dedicated infrastructure | — | — | — | — | ✓ |
| SSO | — | — | — | — | ✓ |
Built for production.
Keys, history, retention, and auditability that already ship.
API authentication
Calls use a bearer key. Live keys and test keys are separate, and a secret is shown once.
Encrypted secrets
A provider key is stored with AES-256-GCM. The dashboard shows a hint, not the secret.
Usage controls
Decision executions are rate limited per key.
Decision history
Each execution is logged with its decision, version, status, and selected option.
Auditability
Key, credential, membership, and publish changes are written to an audit log.
Configurable retention
Context is kept for the retention window on that decision, then removed.
Decisions, keys, and overage.
What is a decision execution?
Every call to owctopus.decide() counts as one decision execution. It counts when a condition answers, when AI answers, when a pattern or a score is used, or when those are combined. You pay for the decision, not for which strategy ran inside it.
Do I need to provide my own AI key?
Yes. Connect your own provider key. Owctopus uses it when a decision is still unknown. OpenAI-compatible keys are supported, and the provider list can grow.
Does Owctopus charge for AI calls?
With your own key, model usage is billed by your AI provider. Owctopus charges for decision infrastructure: the execution, the history, and the conditions learned from it.
Does Owctopus reduce AI usage?
Owctopus is designed to recognize repeated patterns and handle known decisions with conditions, so those requests do not need another model call. How far that goes depends on how often your decisions repeat. There is no guaranteed percentage.
Where are my conditions stored?
Owctopus stores and manages conditions as part of your decision infrastructure. Your application keeps sending decide(). It does not have to host the conditions itself.
Can I use OpenAI?
Yes. Connect an OpenAI-compatible key. The call stays on your provider bill.
What currency is pricing in?
List prices are shown in USD by default. Visitors in India see INR. You can switch anytime. AI cost estimates use published USD model rates and convert for display when you choose INR.
What happens if I exceed my plan?
Developer, Growth, and Scale add an overage price for each decision past the included amount. A monthly spending limit is how you cap that overage so the bill stays predictable. Past 10 million decisions, the fit is Enterprise.
Can I upgrade later?
Yes. Move to a larger plan when the volume, or the learning features, call for it. The decide() call in your application stays the same.
Can I cancel?
Yes. You can cancel a paid plan. The Free plan has no monthly charge.
Install the SDK. Connect your AI. Measure one workload.
Start free. Use the calculator with your real model costs before you scale.