titan
GPU compute API: LLM inference (ornith-1.5) on RTX 4080, pay-per-credit in Bitcoin.
我该使用它吗
质量与安全性
A
基于对工具定义和协议合规性的自动分析。
上下文开销
~336token 数（工具定义）
~427 B典型响应大小
对注意力的影响极小（占 128k 上下文窗口的 0.26%）
这是每次将服务器的工具加载到模型上下文窗口时所消耗的大致 token 数。数值越高，可用于其他任务的注意力就越少。
安装
一键安装
将以下内容添加到你的 `claude_desktop_config.json` 文件中：
{
"mcpServers": {
"titan": {
"url": "https://titan-mcp.thetempleofdoom.com/mcp"
}
}
}远程端点
https://titan-mcp.thetempleofdoom.com/mcpstreamable-http它能做什么
工具清单
工具（5）
🟢 只读🟡 写入🔴 删除⚪ 未知
🟢titan_beacon
Check TITAN GPU compute status (free, no key needed).
输入模式
{
"type": "object",
"properties": {},
"title": "titan_beaconArguments"
}⚪titan_generate(prompt, model, api_key)
Run LLM inference on the RTX 4080. Costs 1 credit. Requires an API key.
输入模式
{
"type": "object",
"properties": {
"prompt": {
"title": "Prompt",
"type": "string"
},
"model": {
"default": "qwen3.8fast",
"title": "Model",
"type": "string"
},
"api_key": {
"default": "",
"title": "Api Key",
"type": "string"
}
},
"required": [
"prompt"
],
"title": "titan_generateArguments"
}🟡titan_order(credits)
Create a payment order for TITAN credits (returns a BTCPay checkout URL).
输入模式
{
"type": "object",
"properties": {
"credits": {
"default": 100,
"title": "Credits",
"type": "integer"
}
},
"title": "titan_orderArguments"
}⚪titan_collect_key(order_id)
Collect your TITAN API key after paying (status=settled).
输入模式
{
"type": "object",
"properties": {
"order_id": {
"title": "Order Id",
"type": "string"
}
},
"required": [
"order_id"
],
"title": "titan_collect_keyArguments"
}🟢titan_credits(api_key)
Check remaining credits for an API key.
输入模式
{
"type": "object",
"properties": {
"api_key": {
"default": "",
"title": "Api Key",
"type": "string"
}
},
"title": "titan_creditsArguments"
}社区
证据
最近观测
已验证未记录版本5 个工具