Generate AI video, images and music, chat with LLMs, and manage a prepaid credits wallet.
Maintain this server? Claim it
Call this server's tools from your browser. Nothing to install.
Optional. "Bearer tmx_..." personal API key from https://turbomaxai.com/developers. Catalogue tools work without it; wallet and generation tools need it.
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.
Work other users chose to publish, with the prompt and model used. Public; anonymous.
1-50, default 12.
Skip this many (paging).
Pick your client. Values in capitals, like YOUR_TOKEN, are yours to fill in.
claude mcp add --transport http mcp https://turbomaxai.com/mcp
Headers
Authorization (optional, not in the config above): Optional. "Bearer tmx_..." personal API key from https://turbomaxai.com/developers. Catalogue tools work without it; wallet and generation tools need it.Work other users chose to publish, with the prompt and model used. Public; anonymous.
| Argument | Type | Description |
|---|---|---|
| limit | integer | 1-50, default 12. |
| offset | integer | Skip this many (paging). |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"maximum": 50,
"minimum": 1,
"description": "1-50, default 12."
},
"offset": {
"type": "integer",
"minimum": 0,
"description": "Skip this many (paging)."
}
},
"additionalProperties": false
}Stop a job that is still processing and refund its hold. Requires a "generate" key.
| Argument | Type | Description |
|---|---|---|
| taskId* | string | The taskId to cancel. |
{
"type": "object",
"required": [
"taskId"
],
"properties": {
"taskId": {
"type": "string",
"description": "The taskId to cancel."
}
},
"additionalProperties": false
}Send one chat turn to a TurboMax text model and return its reply. SPENDS CREDITS (held, then settled to actual usage). Stateless: send the whole conversation in messages, or just message for a single question. Requires a "generate" key.
| Argument | Type | Description |
|---|---|---|
| message | string | A single user message. |
| modelId | string | Chat model id from list_chat_models. Omit for the default. |
| messages | array | The conversation so far; the last message must be from the user. |
| reasoningEffort | string | Only for models that support it. |
{
"type": "object",
"properties": {
"message": {
"type": "string",
"description": "A single user message."
},
"modelId": {
"type": "string",
"description": "Chat model id from list_chat_models. Omit for the default."
},
"messages": {
"type": "array",
"items": {
"type": "object",
"required": [
"role",
"content"
],
"properties": {
"role": {
"enum": [
"user",
"assistant"
],
"type": "string"
},
"content": {
"type": "string"
}
},
"additionalProperties": false
},
"description": "The conversation so far; the last message must be from the user."
},
"reasoningEffort": {
"enum": [
"low",
"medium",
"high"
],
"type": "string",
"description": "Only for models that support it."
}
},
"additionalProperties": false
}Approximate credits for a media generation from the live price list. The exact amount is held when you call generate_media; pass maxCredits there to make it a hard ceiling.
| Argument | Type | Description |
|---|---|---|
| modelId* | string | Model id from list_models. |
| duration | number | Seconds, for video models. |
| resolution | string | e.g. "720p", "1080p". Omit for the model default. |
{
"type": "object",
"required": [
"modelId"
],
"properties": {
"modelId": {
"type": "string",
"description": "Model id from list_models."
},
"duration": {
"type": "number",
"description": "Seconds, for video models."
},
"resolution": {
"type": "string",
"description": "e.g. \"720p\", \"1080p\". Omit for the model default."
}
},
"additionalProperties": false
}Start a generation. SPENDS CREDITS from the wallet (held now, settled to the actual cost when the job ends; refunded if it fails). Returns a taskId immediately - follow it with get_task (waitSeconds) until completed. Use only the user's own prompt text. For image-to-video etc. upload the source first (upload_source_media). Requires an API key with the "generate" scope.
| Argument | Type | Description |
|---|---|---|
| mode | string | From the model's modes, if it has any. |
| task | string | Task, e.g. "text-to-video", "image-to-video". Must be one the model supports. |
| ratio | string | Aspect ratio from the model's aspectRatios, e.g. "16:9". |
| prompt* | string | What to create, in the user's words. |
| modelId* | string | Model id from list_models. |
| duration | number | Seconds, within the model's durationRange (respect its step). |
| maxCredits | number | Spending ceiling. If the cost would exceed it the server refuses and charges nothing. Always set this when no human is watching. |
| resolution | string | From the model's resolutions, e.g. "720p". |
Full parameter details and pricing for one media model.
| Argument | Type | Description |
|---|---|---|
| modelId* | string | Model id from list_models. |
{
"type": "object",
"required": [
"modelId"
],
"properties": {
"modelId": {
"type": "string",
"description": "Model id from list_models."
}
},
"additionalProperties": false
}Status, progress and result of a job started with generate_media. With waitSeconds it waits for completion. resultUrl is a time-limited link - download or use it promptly. Polling also settles the wallet when the job ends.
| Argument | Type | Description |
|---|---|---|
| taskId* | string | The taskId from generate_media (starts with "tbmx_"). |
| waitSeconds | integer | Wait up to this many seconds for a final state (0-55). Default 0. |
{
"type": "object",
"required": [
"taskId"
],
"properties": {
"taskId": {
"type": "string",
"description": "The taskId from generate_media (starts with \"tbmx_\")."
},
"waitSeconds": {
"type": "integer",
"maximum": 55,
"minimum": 0,
"description": "Wait up to this many seconds for a final state (0-55). Default 0."
}
},
"additionalProperties": false
}Credit balance, credits-per-dollar, today's usage against the account limits, and recent ledger entries.
| Argument | Type | Description |
|---|---|---|
| transactions | integer | How many recent ledger entries to include (0-50, default 10). |
{
"type": "object",
"properties": {
"transactions": {
"type": "integer",
"maximum": 50,
"minimum": 0,
"description": "How many recent ledger entries to include (0-50, default 10)."
}
},
"additionalProperties": false
}The text/chat models available to chat, with the default and whether each is currently available.
No arguments.
{
"type": "object",
"properties": {},
"additionalProperties": false
}The account's generation history, newest first, with status, cost and result links.
| Argument | Type | Description |
|---|---|---|
| limit | integer | 1-50, default 12. |
| offset | integer | Skip this many (paging). |
| status | string |
{
"type": "object",
"properties": {
"limit": {
"type": "integer",
"maximum": 50,
"minimum": 1,
"description": "1-50, default 12."
},
"offset": {
"type": "integer",
"minimum": 0,
"description": "Skip this many (paging)."
},
"status": {
"enum": [
"processing",
"completed",
"failed",
"canceled"
],
"type": "string"
}
},
"additionalProperties": false
}List the image, video and audio models with what each accepts (tasks, aspect ratios, resolutions, duration range, modes) and its price in credits. Call this before generate_media; use the exact ids and values it returns.
| Argument | Type | Description |
|---|---|---|
| task | string | Only models supporting this task, e.g. "text-to-video", "image-to-video", "text-to-image", "image-editing", "text-to-music". |
| type | string | Only models of this media type. |
{
"type": "object",
"properties": {
"task": {
"type": "string",
"description": "Only models supporting this task, e.g. \"text-to-video\", \"image-to-video\", \"text-to-image\", \"image-editing\", \"text-to-music\"."
},
"type": {
"enum": [
"video",
"image",
"audio"
],
"type": "string",
"description": "Only models of this media type."
}
},
"additionalProperties": false
}Re-run a failed or cancelled job as a new job. SPENDS CREDITS again (a fresh hold). Requires a "generate" key.
| Argument | Type | Description |
|---|---|---|
| taskId* | string | The failed/cancelled taskId. |
{
"type": "object",
"required": [
"taskId"
],
"properties": {
"taskId": {
"type": "string",
"description": "The failed/cancelled taskId."
}
},
"additionalProperties": false
}Upload an image, video or audio file (PNG, JPEG, WebP, GIF, MP4, WebM, MOV, MP3, WAV) to use as sourceImageUrl in generate_media. Send the bytes as base64 (or a local filePath when this server runs locally). Requires a "generate" key.
| Argument | Type | Description |
|---|---|---|
| filePath | string | Absolute path to a local file. Only available when the MCP server runs on your own machine (stdio). |
| filename | string | Original file name (informational). |
| dataBase64 | string | The file bytes, base64-encoded (no data: prefix). |
{
"type": "object",
"properties": {
"filePath": {
"type": "string",
"description": "Absolute path to a local file. Only available when the MCP server runs on your own machine (stdio)."
},
"filename": {
"type": "string",
"description": "Original file name (informational)."
},
"dataBase64": {
"type": "string",
"description": "The file bytes, base64-encoded (no data: prefix)."
}
},
"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 2, 2026.
Asks the model to put the whole conversation, its system prompt or other tools' output into an argument: a data exfiltration pattern.
Fix: Ask only for the inputs the tool needs, in fields the user can see.
| stylePreset | string | From the model's stylePresets. Omit for none (the prompt is used verbatim). |
| waitSeconds | integer | Wait up to this many seconds for the result before returning (0-55). Default 0: return the taskId at once. |
| cameraMotion | string | From the model's cameraMotions. |
| generateAudio | boolean | For models with an audio toggle. |
| isInstrumental | boolean | Audio models: no vocals. |
| negativePrompt | string | What to avoid (only models with hasNegativePrompt). |
| sourceImageUrl | string | URL returned by upload_source_media. Only uploads made through TurboMax are accepted. |
{
"type": "object",
"required": [
"modelId",
"prompt"
],
"properties": {
"mode": {
"type": "string",
"description": "From the model's modes, if it has any."
},
"task": {
"type": "string",
"description": "Task, e.g. \"text-to-video\", \"image-to-video\". Must be one the model supports."
},
"ratio": {
"type": "string",
"description": "Aspect ratio from the model's aspectRatios, e.g. \"16:9\"."
},
"prompt": {
"type": "string",
"description": "What to create, in the user's words."
},
"modelId": {
"type": "string",
"description": "Model id from list_models."
},
"duration": {
"type": "number",
"description": "Seconds, within the model's durationRange (respect its step)."
},
"maxCredits": {
"type": "number",
"minimum": 0,
"description": "Spending ceiling. If the cost would exceed it the server refuses and charges nothing. Always set this when no human is watching."
},
"resolution": {
"type": "string",
"description": "From the model's resolutions, e.g. \"720p\"."
},
"stylePreset": {
"type": "string",
"description": "From the model's stylePresets. Omit for none (the prompt is used verbatim)."
},
"waitSeconds": {
"type": "integer",
"maximum": 55,
"minimum": 0,
"description": "Wait up to this many seconds for the result before returning (0-55). Default 0: return the taskId at once."
},
"cameraMotion": {
"type": "string",
"description": "From the model's cameraMotions."
},
"generateAudio": {
"type": "boolean",
"description": "For models with an audio toggle."
},
"isInstrumental": {
"type": "boolean",
"description": "Audio models: no vocals."
},
"negativePrompt": {
"type": "string",
"description": "What to avoid (only models with hasNegativePrompt)."
},
"sourceImageUrl": {
"type": "string",
"description": "URL returned by upload_source_media. Only uploads made through TurboMax are accepted."
}
},
"additionalProperties": false
}