
Search and read the KovAPI API docs as Markdown and list its models with prices. Read-only, no key.
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.
Every model on sale at KovAPI with its id, name, vendor, category and USD price, plus the link to its API doc (Markdown).
Only this category.
This tool isn't marked read-only, so it may change something on the server or in an account you connected.
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
Run in a terminal
claude mcp add --transport http docs https://kovapi.com/mcp
Every model on sale at KovAPI with its id, name, vendor, category and USD price, plus the link to its API doc (Markdown).
| Argument | Type | Description |
|---|---|---|
| category | string | Only this category. |
{
"type": "object",
"properties": {
"category": {
"enum": [
"chat",
"image",
"video"
],
"type": "string",
"description": "Only this category."
}
}
}Return one KovAPI documentation page as Markdown. Pass a path or URL, e.g. /docs/api, /docs/nano-banana-2, /docs/guide/errors, /docs/integrations/claude-code, /docs-guide, /docs/webhooks, or a docs.kovapi.com URL. A language prefix (/zh /en /es /ja /ko /ru) picks the language; the bare path is the site default.
| Argument | Type | Description |
|---|---|---|
| path* | string | Docs path or full URL. |
{
"type": "object",
"required": [
"path"
],
"properties": {
"path": {
"type": "string",
"description": "Docs path or full URL."
}
}
}Search the KovAPI API documentation: guide pages and sections, every model doc, endpoints, integrations and the FAQ. Returns titles with URLs; read one with read_doc.
| Argument | Type | Description |
|---|---|---|
| lang | string | Language of the docs to search (default en). |
| query* | string | Words to look for, e.g. "image to image", "webhook signature", "429". |
{
"type": "object",
"required": [
"query"
],
"properties": {
"lang": {
"enum": [
"zh",
"en",
"es",
"ja",
"ko",
"ru"
],
"type": "string",
"description": "Language of the docs to search (default en)."
},
"query": {
"type": "string",
"description": "Words to look for, e.g. \"image to image\", \"webhook signature\", \"429\"."
}
}
}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 21, 2026.
No readOnlyHint / destructiveHint annotations, so clients can't tell a read from a write without asking.
Fix: Add annotations to each tool, e.g. { readOnlyHint: true } for tools that only read.