Pricing
AI model costs included, built for modern AI teams
Your subscription includes AI model calls via OpenRouter, pick any model, pay one flat rate
Compare plans
Free
Explore & experiment
~250 AI model calls / month included
Gemini Flash Lite / GPT-5 mini only
- 1 user
- 10 prompts
- 3 datasets
- 10 eval runs / month
- Version history: 30 days
- Community support
- BYOK support (no credit deduction)
Starter
Solo PM or developer
~35% off with annual billing (₩32,000 / mo)
~4,400 AI model calls / month included
Based on Gemini 3 Flash (varies by model) · Unlimited calls with BYOK
- 1 user (solo)
- 50 prompts
- 10 datasets
- 100 eval runs / month
- Version history: 90 days
- Korean eval: all 7 criteria
- Environment-based deploy (staging/prod)
- 1 alert / 5 API keys
- Change log: 30 days
- Email support (48 h)
Team
3-person PM & dev team
~35% off with annual billing (₩65,000 / mo)
~8,800 AI model calls / month included
Based on Gemini 3 Flash (varies by model) · Shared across 3 seats, +₩30,000/extra seat · Unlimited calls with BYOK
- 3 users included
- 200 prompts
- 100 datasets
- 500 eval runs / month
- Version history: 365 days
- Korean eval: 7 criteria + custom
- Compare 3+ models simultaneously
- Custom domain (share pages)
- 5 alerts / 10 API keys
- Change log: 90 days
- Email support (24 h)
Why PromptsOps costs less
Most LLMOps tools charge extra for AI model usage · PromptOps bundles it in
| Tool | Monthly fee | AI model costs | True cost |
|---|---|---|---|
| Langfuse Cloud Pro | $59 / mo | Separate (BYOK) | $59 + your AI model costs |
| LangSmith Plus | $39 / mo | Separate (BYOK) | $39 + your AI model costs |
| Helicone Pro | $20 + per-req fee | Separate (BYOK) | $20+ + your AI model costs |
| PromptsOps Starter | ₩49,000 | Included | ₩49,000 (AI model calls included) |
Detailed plan comparison
| Feature | Free | Starter | Team |
|---|---|---|---|
| Monthly price | ₩0 | ₩49,000 | ₩99,000 |
| Extra users | – | – | +₩30,000 / seat |
| AI model calls included (Gemini Flash basis) | ~250 calls | ~4,400 calls | ~8,800 calls (3 seats shared) |
| Available models | Lite / Mini only | All (excl. gpt-5.5-pro) | All (excl. gpt-5.5-pro) |
| BYOK (bring your own API key) | (no credit deduction) | Included | Included |
| Prompts | 10 | 50 | 200 |
| Datasets | 3 | 10 | 100 |
| Eval runs | 10 / mo | 100 / mo | 500 / mo |
| Version history | 30 days | 90 days | 365 days |
| Korean response eval | 3 basic criteria | All 7 criteria | 7 criteria + custom |
| Environment-based deploy (staging/prod) | – | Included | Included |
| Compare 3+ models simultaneously | – | – | Included |
| Custom domain (share pages) | – | – | Included |
| Change log retention | 7 days | 30 days | 90 days |
| API keys / alerts | 1 / – | 5 / 1 | 10 / 5 |
| Support SLA | Community | Email 48 h | Email 24 h |
AI model call counts are estimates based on Gemini 3 Flash (a cost-efficient model), using premium models (GPT-5.1, Claude Sonnet, Opus, etc.) automatically adjusts available calls. When credits run out, register your own API key (BYOK) or set a top-up limit billed at 1.5× OpenRouter list price
Frequently asked questions
Still have questions? Email us at support@promptsops.com
What is PromptOps?
Are AI model call costs included in the subscription?
What happens when I run out of credits?
Can I keep using my existing OpenAI or Anthropic SDK code?
How does Korean language evaluation work?
What is the payment and refund policy?
Get started today
No credit card required
Start for free