Protect profit before usage scales
Growth can hide unprofitable AI customers.
More customers and more usage should create more value—not larger invisible losses. Ganivra shows which customers, workflows, and agents make money, where margin is disappearing, and what to fix before you scale. From $29 per month.
For early-stage AI products
Starter
See what remains of customer revenue after recorded costs.
- Track multiple applications in one workspace
- Unlimited customer IDs
- 90-day economics history
- Customer and workflow revenue, cost, and contribution
- Agent cost drivers with model and MCP execution detail
- Versioned provider pricing and private rates
- Margin-risk alerts
For startups finding traction
Launch
Find customers and workflows whose usage is outgrowing their revenue.
- Everything in Starter
- 2.5× the included call volume
- 90-day economics history
- Unlimited customer IDs
For scaling AI products
Growth
Understand changing margins and decide where to adjust pricing or costs.
- Everything in Launch
- 12-month economics history
- Cohorts, trends, drivers, and pricing headroom
- Tool fan-out and reported retry cost trends
- Revenue and billing-actual imports
- Historical repricing
- Provider-invoice reconciliation
For high-volume AI products
Scale
Keep customer and workflow economics visible as adoption grows.
- Everything in Growth
- 3× the included call volume
- 12-month economics history
- Unlimited customer IDs
Enterprise · Custom pricing
Need more than 300,000 calls?
Talk to us about custom volume, retention, implementation support, and commercial terms.
Why teams invest
More tokens. More customers. Where is the margin?
Your AI service is getting used. But longer conversations, more agent steps, and retries can grow costs faster than revenue. A customer paying $50 can cost $80 to serve. Ganivra shows the gap and the workflows behind it, so you can adjust pricing or delivery before more adoption means more losses.
- Subscription revenue
- $50
- Recorded AI + delivery costs
- −$80
- Contribution after recorded costs
- −$30
Contribution is revenue minus recorded costs, not net profit. Connect revenue and report relevant delivery costs for a meaningful picture.
How usage is counted
One shared allowance for model and tool calls.
Accepted model and MCP events
Each accepted model or MCP tool-call event counts once, including failed calls and retries with distinct event IDs. Your workspaces share the plan allowance.
Everything around the call
Duplicates, rejected events, revenue entries, repricing, dashboard reads, alerts, and configuration are not billable calls.
Customer attribution
Pass as many B2B tenant or account IDs as your product serves. The value of unit economics depends on seeing all of them.
Pricing questions
Know exactly what you are buying.
01What business question does Ganivra answer?
Are your customers and AI workflows making money after recorded costs? Ganivra connects customer revenue to model, tool, and reported delivery costs, shows contribution by customer and workflow, and identifies agent cost drivers. Use the evidence to adjust pricing, allowances, or expensive workflows.
02Can we use Ganivra alongside our observability tools?
Yes. Ganivra focuses on the economics of delivering your product: customer revenue, cost-to-serve, workflow contribution, and agent cost drivers. Keep your existing debugging and evaluation tools, and use Ganivra to understand the financial impact of the usage you record.
03What do we need to see reliable margins?
Send consistent customer, workflow, and agent attribution. Connect revenue through configured customer plans or confirmed revenue records, and report delivery costs such as telephony, OCR, and compute separately. Missing revenue and unpriced usage stay visible. Contribution after recorded costs does not imply every business expense is covered.
04Which plan fits our monthly usage?
Starter includes 10,000 accepted AI calls for $29 per month. Launch includes 25,000 for $49, Growth includes 100,000 for $99, and Scale includes 300,000 for $199. Choose the allowance that fits your expected monthly volume. Contact us for higher volumes or custom requirements.
05Does Ganivra track MCP tools?
Yes. Send MCP steps through the same event API as model calls, identify the server and tool, and keep every step on the same execution ID. Ganivra can show the direct tool cost plus instrumented planning, retrieval, validation, retry, and model steps. Each actual retry must be reported with its own event; internal provider SDK retries may not be separately visible.
06What is an accepted AI call?
One valid model-call or MCP tool-call event accepted by Ganivra. Both share the plan allowance across your primary workspace and its additional workspaces. Calls can come from OpenAI, Google Gemini, Anthropic Claude, Perplexity Sonar, or another provider through the REST event contract.
07What does not count toward usage?
Rejected events, duplicate event IDs within the same workspace, revenue-ledger entries, historical repricing, dashboard views, and pricing or customer-plan configuration do not count as AI calls.
08Do customer IDs increase the price?
No. Every paid plan supports unlimited customer IDs. A B2B AI company can send its existing tenant or account ID dynamically on each event and analyze every customer's cost-to-serve without registering them manually.
09Do we have to send revenue on every call?
No. Configure and assign a subscription, token, credit, per-execution, or hybrid customer plan in Ganivra. Your application sends the customer and usage facts—not prices. Credit plans send the units deducted; the other plan types use captured execution and token telemetry.
Built for ambitious AI companies
Scale what makes money—not what quietly destroys margin.
See where token spend is outgrowing customer revenue. Find the customers, workflows, and agents to fix before growth makes the problem larger.