Skip to main content
Glama
deenesjakoruzh
newageflyfish-max

volthq-mcp-server

by newageflyfish-max
OverviewSchemaSourceRelated ServersDiscussionsAdmin
TypeScript
Remote

Volt HQ

AI 智能体的计算价格预言机。

功能

  • 比较定价：涵盖 9 家提供商（OpenAI、Anthropic、Groq、Together AI、DeepInfra、Fireworks AI、Cerebras、Hyperbolic、Akash）—— 200 多种产品，提供实时 API 定价

  • 推荐最优路由：告诉您的智能体在哪里可以用更低的价格获得相同的质量，并提供节省估算

  • 跟踪支出和预算：按提供商/模型分类的支出摘要、节省报告和阈值提醒

Related MCP server: volthq-mcp-server

安装

通过一条命令自动配置 Cursor 和 Claude Desktop：

npx volthq-mcp-server --setup

检测已安装的客户端，合并配置而不覆盖您现有的 MCP 服务器。

Cursor — 添加到 .cursor/mcp.json：

{
  "mcpServers": {
    "volthq": {
      "command": "npx",
      "args": ["-y", "volthq-mcp-server"]
    }
  }
}

Claude Desktop — 添加到 claude_desktop_config.json：

{
  "mcpServers": {
    "volthq": {
      "command": "npx",
      "args": ["-y", "volthq-mcp-server"]
    }
  }
}

工具

工具

描述

volt_check_price

比较特定模型在各提供商之间的定价

volt_recommend_route

获取最优提供商建议及节省估算

volt_get_spend

按提供商和模型分类的支出摘要（今日/7天/30天）

volt_get_savings

实际支出与优化后支出的对比

volt_set_budget_alert

设置每日/每周/每月的预算阈值提醒

示例

> volt_check_price { "model": "llama-70b" }

Price comparison for "llama-70b" — 14 offerings found
────────────────────────────────────────────────────────────
1. DeepInfra — Llama-70B
   Input: $0.20/M tokens | Output: $0.27/M tokens | Avg: $0.24/M
   Quality: 88% | Region: global

2. Hyperbolic — Llama-70B (FP8) on H100-SXM
   Input: $0.40/M tokens | Output: $0.40/M tokens | Avg: $0.40/M
   Quality: 85% | Region: global

3. Hyperbolic — Llama-70B (BF16) on H100-SXM
   Input: $0.55/M tokens | Output: $0.55/M tokens | Avg: $0.55/M
   Quality: 88% | Region: global

4. Groq — Llama-70B
   Input: $0.59/M tokens | Output: $0.79/M tokens | Avg: $0.69/M
   Quality: 88% | Region: global

5. Fireworks AI — Llama-70B
   Input: $0.90/M tokens | Output: $0.90/M tokens | Avg: $0.90/M
   Quality: 88% | Region: global

6. Together AI — Llama-70B
   Input: $0.88/M tokens | Output: $0.88/M tokens | Avg: $0.88/M
   Quality: 88% | Region: global

7. Akash — Llama-70B (FP8) on H100-SXM
   Input: $3.49/M tokens | Output: $8.72/M tokens | Avg: $6.11/M
   Quality: 85% | Region: global

8. Akash — Llama-70B (FP8) on A100-80GB
   Input: $5.24/M tokens | Output: $13.11/M tokens | Avg: $9.18/M
   Quality: 85% | Region: global

Cheapest is 97% less than most expensive option.

DeepInfra 为 $0.24/M，Hyperbolic 为 $0.40/M，Groq 为 $0.69/M，Fireworks AI 为 $0.90/M —— 相比之下 GPT-4o 为 $6.25/M。

支持的提供商

  • OpenAI — GPT-4o, GPT-4o-mini

  • Anthropic — Claude Sonnet 4.6, Claude Haiku 4.5

  • Groq — Llama-70B, Llama-8B, Mixtral-8x7B

  • Together AI — Llama-70B, Llama-8B, DeepSeek-V3

  • DeepInfra — 75+ 种具有实时 API 定价的模型（Llama、DeepSeek、Qwen、Mistral、Gemma 等）

  • Fireworks AI — Llama-70B, Llama-8B, DeepSeek-V3

  • Cerebras — Llama-3.1-8B, Qwen-3-235B

  • Hyperbolic — DeepSeek-V3, DeepSeek-R1, Llama-70B, Llama-8B

  • Akash — 基于 H100 和 A100 的 Llama-70B, Llama-8B（实时 GPU 定价）

遥测

Volt 默认收集匿名使用元数据以改进路由建议。这包括：提供商名称、模型名称、工具响应时间和成功/失败状态。

绝不收集的内容： 提示词、输出、API 密钥、Token 计数或任何用户可识别的内容。IP 地址在服务器端进行哈希处理和截断。

如需退出，请设置环境变量：

VOLT_OBSERVATIONS=false

链接

许可证

MIT

This server cannot be deployed

One endpoint for every MCP toolRayrun runs every MCP server behind one URL and holds the credentials. Each call is decided per client, tool, and argument, then recorded.ray.runAd

Maintenance

ActivityInactive
ResponsivenessNo issues

Resources

Related MCP Connectors

Related MCP Servers

  • AgentCost MCP Server

    oneagentlabs
    F
    license
    Not graded
    quality
    D
    maintenance
    Provides real-time AI model pricing, cost estimation, and budget management tools to help agents understand and optimize their spending. It enables agents to compare costs across multiple providers and select the most cost-effective models for specific tasks.
    1
    -

  • atom-mcp-server

    A7OM-AI
    A
    license
    A
    quality
    C
    maintenance
    Global price benchmarking for AI inference across 2,600+ SKUs from 47 vendors. Query live pricing, market indexes, and model specs via 8 tools. Free tier available.
    8
    53 npm
    MIT