Estimate AI service costs, setup fees, monthly retainers, overages, and gross margin.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Requests go through mcptry because browsers can't call most MCP servers directly. Header values are sent to the server for that request only and are never stored or logged.
Tool list from our last check. Connect live to refresh it.
Estimate AI service costs, retainer, setup fee, overages and margin in USD. User-supplied model rates. No data saved or purchases made.
Your delivery cost ($ / hour)
Input price ($ / million tokens)
Setup + implementation hours
Input tokens / model call
Output price ($ / million tokens)
Output tokens / model call
Hosting + tools ($ / month)
Support hours / month
Target gross margin (%)
Model calls / request
Client requests / month
Cost contingency (%)
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http llm-margin https://llmmargin.com/mcp
Estimate AI service costs, retainer, setup fee, overages and margin in USD. User-supplied model rates. No data saved or purchases made.
| Argument | Type | Description |
|---|---|---|
| model | string | |
| scope | string | |
| title | string | |
| agency | string | |
| client | string | |
| hourlyCost | number | Your delivery cost ($ / hour) |
| inputPrice | number | Input price ($ / million tokens) |
| setupHours | number | Setup + implementation hours |
| inputTokens | number | Input tokens / model call |
| outputPrice | number | Output price ($ / million tokens) |
| outputTokens | number | Output tokens / model call |
| platformCost | number | Hosting + tools ($ / month) |
| supportHours | number | Support hours / month |
| marginPercent | number | Target gross margin (%) |
| callsPerRequest | number | Model calls / request |
| monthlyRequests | number | Client requests / month |
| contingencyPercent | number | Cost contingency (%) |
{
"type": "object",
"properties": {
"model": {
"type": "string",
"maxLength": 120
},
"scope": {
"type": "string",
"maxLength": 4000
},
"title": {
"type": "string",
"maxLength": 120
},
"agency": {
"type": "string",
"maxLength": 120
},
"client": {
"type": "string",
"maxLength": 120
},
"hourlyCost": {
"type": "number",
"default": 60,
"maximum": 10000,
"minimum": 0,
"description": "Your delivery cost ($ / hour)"
},
"inputPrice": {
"type": "number",
"default": 3,
"maximum": 10000,
"minimum": 0,
"description": "Input price ($ / million tokens)"
},
"setupHours": {
"type": "number",
"default": 20,
"maximum": 10000,
"minimum": 0,
"description": "Setup + implementation hours"
},
"inputTokens": {
"type": "number",
"default": 2000,
"maximum": 10000000,
"minimum": 0,
"description": "Input tokens / model call"
},
"outputPrice": {
"type": "number",
"default": 15,
"maximum": 10000,
"minimum": 0,
"description": "Output price ($ / million tokens)"
},
"outputTokens": {
"type": "number",
"default": 500,
"maximum": 1000000,
"minimum": 0,
"description": "Output tokens / model call"
},
"platformCost": {
"type": "number",
"default": 50,
"maximum": 1000000,
"minimum": 0,
"description": "Hosting + tools ($ / month)"
},
"supportHours": {
"type": "number",
"default": 3,
"maximum": 10000,
"minimum": 0,
"description": "Support hours / month"
},
"marginPercent": {
"type": "number",
"default": 60,
"maximum": 95,
"minimum": 0,
"description": "Target gross margin (%)"
},
"callsPerRequest": {
"type": "number",
"default": 2,
"maximum": 1000,
"minimum": 1,
"description": "Model calls / request"
},
"monthlyRequests": {
"type": "number",
"default": 10000,
"maximum": 100000000,
"minimum": 1,
"description": "Client requests / month"
},
"contingencyPercent": {
"type": "number",
"default": 20,
"maximum": 200,
"minimum": 0,
"description": "Cost contingency (%)"
}
},
"additionalProperties": false
}We read every tool's name, description and inputs the way the model will, looking for instructions hidden from you. A clean result means no warning signs, not a guarantee. Scanned Nov 11, 2026.