Transparent plans.
Scale with confidence.
Start on Free. When you need a real feedback loop, Starter at $29.90 ships the essentials. Pro scales to production. Enterprise covers compliance and isolation.
- 50,000 requests/month
- 20 providers (OpenAI, Anthropic, Gemini…)
- Exact cache + basic firewall (LLM Firewall)
- 7-day log retention
- 100k requests/month + 1k rpm rate limit
- Feedback API (500/month)
- Semantic cache
- 90-day log retention
- Feedback-driven routing
- Smart selectors + A/B testing
- Advanced firewall (LLM Firewall)
- 2-year retention + 10k rpm rate limit
- SSO/SAML + SOC-2 + HIPAA
- SLA + dedicated support (Slack + engineer)
- Opt-out of shared model
- Dedicated tenant isolation + custom retention
What would Floopy pay back?
Estimate based on our median deployment. Actual savings depend on traffic mix and quality tolerance — most teams see 40–65%.
Everything, side by side.
| Feature | Free | Starter | Pro | Enterprise |
|---|---|---|---|---|
| Routing & Optimization | ||||
| Feedback-driven routing Routes each request based on prior-session NPS, improving model choice as user feedback accumulates. | — | — | ● | ● + custom model |
| Smart selectors Selectors that pick the right model per task class (reasoning, extraction, summarization) without you hardcoding the rule. | — | — | ● | ● |
| Smart Cost Routing Swaps expensive models for cheaper alternatives when observed quality is equivalent, cutting cost without sacrificing performance. | — | ● | ● | ● |
| A/B testing & experiments | — | ● | ● | ● |
| Regression monitoring | — | — | ● | ● |
| Gateway & Providers | ||||
| Supported providers | 20 | 20 | 20 | 20 |
| OpenAI-compatible API | ● | ● | ● | ● |
| Automatic fallback | ● | ● | ● | ● |
| Rate Limit (RPS / 10s burst / RPM) | 5 / 50 / 100 | 30 / 300 / 1000 | 100 / 1,000 / 10,000 | custom |
| Cache & Performance | ||||
| Exact cache | ● | ● | ● | ● |
| Semantic cache | — | ● | ● | ● |
| Embeddings cache | — | ● | ● | ● |
| LLM Firewall & Security | ||||
| LLM Firewall | ● | ● | ● | ● |
| LLM Firewall (advanced) | — | — | ● | ● |
| Custom guardrails | — | ● | ● | ● |
| Prompts & Datasets | ||||
| Versioned prompts | ● | ● | ● | ● |
| Playground | ● | ● | ● | ● |
| Prompt experiments | — | ● | ● | ● |
| Optimized dataset creator Builds training/eval datasets from real traffic, automatically deduplicating and prioritizing diverse examples. | — | — | ● | ● |
| Observability | ||||
| Requests included | 50k | 100k | 100k + overage | custom |
| Log retention | 7 days | 90 days | 2 years | custom |
| Alerts & webhooks | — | ● | ● | ● |
| Feedback API | — | 500/mo | unlimited | unlimited |
| Collaboration & Limits | ||||
| Seats | 1 | 3 | 5 (+$10/extra) | unlimited |
| Organizations | 1 | 1 | 3 | unlimited |
| Tiered overage Tiered overage billing: the more requests you make in a month, the lower the price per extra request. | — | ● | ● | ● |
| Enterprise & Compliance | ||||
| SSO/SAML | — | — | — | ● |
| SOC-2 | — | — | — | ● |
| HIPAA | — | — | — | ● |
| Audit logs | — | — | — | ● |
| SLA | — | — | — | ● |
Start saving in under 10 minutes.
Frequently asked questions
The questions we actually get from teams evaluating Floopy.
What happens when I exceed 50,000 requests?
On the Free plan, requests are blocked. On Pro, automatic tiered pricing applies — the more you use, the cheaper per request.
Do I need to install an SDK?
No. Floopy works by changing the baseURL in the OpenAI SDK you already use. One line of code, zero dependencies.
Which AI providers are supported?
20 providers, full list: OpenAI, Anthropic, Google Gemini, Google Vertex AI, AWS Bedrock, Azure OpenAI, DeepSeek, Mistral, xAI, Groq, Cerebras, SambaNova, Together AI, Fireworks AI, Perplexity, Cohere, AI21 Labs, DeepInfra, Nebius, and Novita. All accessible through a single OpenAI-compatible endpoint — point your SDK's baseURL at Floopy and call any model from this list with the same shape. Streaming and tool calls are supported across the list; vision inputs are supported on OpenAI, Anthropic, Gemini, and Bedrock today, and rolling out elsewhere as providers ship the capability.
What about enterprise plans?
Contact us for enterprise needs including compliance, dedicated tenant isolation, and custom SLAs.