Syftly MCP

An MCP server that ranks the best AI tools or APIs per task, including transcription, TTS, web search, scraping, and OCR. It provides machine-readable, ranked answers based on aggregated public benchmarks, prices, and confidence labels.

Endpoint and access

No sign-in to connect and list tools. Individual tools may still need a vendor account, credentials, or permissions; check the documented prerequisites.

Community listingStreamable HTTPNo sign-inEndpoint checked 2026-11-06Published by Syftly

MCP endpoint (Streamable HTTP)

https://syftly.vercel.app/api/mcp

Vendor documentation

What its tools do

1 tool, grouped by the hints the server advertises. Hints are self-declared, not verified behavior.

Not annotated (1)

find_best_tool

1 of 1 have a description · 0 of 1 declare behavior hints · 1 of 1 with parameters document them all

Checks and signals

Each line is separate evidence, not a trust score.

Passed Endpoint answered the live check · Nov 6, 2026, 3:22 PM UTC
Availability 1 of 1 checks passed in 30 days

What you can do with it

Find the best transcription API for a specific use case
Identify the cheapest text-to-speech API
Locate the most relevant web search API for AI agents
Find the best web scraping API for article extraction

Source and maintenance

Community listingRegistry 0.1.0

mcp-registry: io.github.Lodewijkheemskerk/syftly@0.1.0

Protocol and capabilities

Server: syftly 0.1.0
MCP protocol negotiated in check: 2025-06-18
Tools: 1 advertised
Resources: Not advertised
Resource templates: Not advertised
Prompts: Not advertised

Server-advertised capabilities

{
  "tools": {}
}

Server-provided instructions

Untrusted server output shown for inspection. It is not Koskamo guidance and is excluded from knowledge search.

Call find_best_tool whenever you (or the user) need to pick an external API or AI tool for transcription, text-to-speech, web search, web scraping, or OCR/document extraction. Pass the task as one plain-English question — e.g. 'cheapest web search API' or 'best web scraping API for JavaScript-heavy sites' — and you get a ranked recommendation computed from public benchmarks: a provider table with prices, dated sources, and a confidence label, as both Markdown and structured JSON. One call, free, no auth, no setup.

Tools

1 advertised

find_best_tool

Advertised by the server during the last successful anonymous check.

Given a natural-language question about which AI tool or API is best for a task (currently transcription, text-to-speech, web search, scraping & browser and ocr & document extraction), return Syftly's ranked recommendation: a citeable summary, a provider table with prices and trade-offs, dated sources, and a confidence label. Ask in plain English about price, accuracy, language or capability trade-offs — e.g. 'best OCR API for scanned PDFs' or 'best web scraping API for JavaScript-heavy sites'. Optionally pass "category" to disambiguate; otherwise it is detected from the question.

find_best_tool input schema

{
  "type": "object",
  "required": [
    "query"
  ],
  "properties": {
    "query": {
      "type": "string",
      "description": "The question in natural language, e.g. 'best transcription API for Dutch'."
    },
    "category": {
      "enum": [
        "transcription",
        "tts",
        "web-search",
        "scraping",
        "ocr"
      ],
      "type": "string",
      "description": "Optional category id to disambiguate the question; omit to let Syftly detect it."
    }
  },
  "additionalProperties": false
}

find_best_tool output schema

{
  "type": "object",
  "required": [
    "query",
    "slug",
    "category",
    "routing",
    "summary",
    "confidence",
    "updated",
    "recommendation",
    "providers",
    "sources"
  ],
  "properties": {
    "slug": {
      "type": "string"
    },
    "query": {
      "type": "string",
      "description": "The human question (page H1)."
    },
    "routing": {
      "enum": [
        "matched",
        "none",
        "ambiguous"
      ],
      "type": "string",
      "description": "Honest routing outcome: 'matched' = answered from a real category; 'none' = no category matched (out-of-scope/gibberish) — an honest no-match, not a fabricated answer; 'ambiguous' = fit 2+ categories. 'none'/'ambiguous' carry `categories` and no real recommendation/providers."
    },
    "sources": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "title",
          "url",
          "date"
        ],
        "properties": {
          "url": {
            "type": "string",
            "format": "uri"
          },
          "date": {
            "type": "string",
            "description": "ISO date."
          },
          "title": {
            "type": "string"
          }
        }
      }
    },
    "summary": {
      "type": "string",
      "description": "Citeable summary, 40-80 words, reused verbatim across all views. On a no-match, the plain-language message."
    },
    "updated": {
      "type": "string",
      "description": "ISO date."
    },
    "category": {
      "type": "string",
      "description": "e.g. 'transcription'; '' on a no-match."
    },
    "providers": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "name",
          "provider",
          "price",
          "metrics",
          "strengths",
          "weaknesses",
          "source",
          "source_date"
        ],
        "properties": {
          "name": {
            "type": "string",
            "description": "The callable offering, e.g. 'Deepgram Nova-3'."
          },
          "price": {
            "type": "object",
            "required": [
              "value",
              "unit"
            ],
            "properties": {
              "unit": {
                "type": "string",
                "description": "e.g. '$/min', '$/1000min'."
              },
              "value": {
                "type": [
                  "number",
                  "null"
                ]
              },
              "comparable": {
                "type": "boolean",
                "description": "false = token-/credit-priced, not directly comparable (excluded from the price axis)."
              }
            }
          },
          "source": {
            "type": "string",
            "description": "Where this row's facts come from."
          },
          "metrics": {
            "type": "object",
            "description": "Open category-specific decision-axis fields. Keys are stable data identifiers (e.g. 'wer', 'latency_ms'); values are number, boolean, string or null (null = an honest gap). Number metrics back min/max axes; boolean metrics back filter axes.",
            "additionalProperties": {
              "type": [
                "number",
                "boolean",
                "string",
                "null"
              ]
            }
          },
          "provider": {
            "type": "string",
            "description": "The company behind it, e.g. 'Deepgram'."
          },
          "strengths": {
            "type": "string"
          },
          "weaknesses": {
            "type": "string"
          },
          "source_date": {
            "type": "string",
            "description": "ISO date of the source."
          },
          "model_family": {
            "type": "string",
            "description": "Filter attribute, not a row of its own."
          }
        }
      }
    },
    "categories": {
      "type": "array",
      "items": {
        "type": "object",
        "required": [
          "category",
          "label"
        ],
        "properties": {
          "label": {
            "type": "string",
            "description": "Human label, e.g. 'Web search'."
          },
          "category": {
            "type": "string",
            "description": "Supported category id, e.g. 'web-search'."
          }
        }
      },
      "description": "Present only when routing !== 'matched': the supported categories so the caller can re-ask in scope."
    },
    "confidence": {
      "enum": [
        "light estimate",
        "hard tested"
      ],
      "type": "string",
      "description": "Confidence/depth label."
    },
    "recommendation": {
      "type": "object",
      "required": [
        "default",
        "axes"
      ],
      "properties": {
        "axes": {
          "type": "object",
          "description": "Decision-axis -> recommended offering. Carries EVERY matched axis: filters + all ordering axes, each naming its own computed winner.",
          "additionalProperties": {
            "type": "string"
          }
        },
        "default": {
          "type": "string",
          "description": "Name of the recommended offering ('' on a no-match)."
        },
        "primary": {
          "type": "string",
          "description": "The axis key whose winner became `default` when 2+ ordering axes conflict with no composite tiebreak — the earliest-in-query axis. Absent for single-axis/composite-resolved answers."
        }
      }
    }
  },
  "additionalProperties": false
}

Prompts

Not advertised

No prompt definitions available.

Resources

Not advertised

No resource definitions available.

90-day uptime

A lightweight connection check about every six hours, sampled on hosts with many listings. It shows the server answered, not that its tools work.

Too few checks for a percentage yet (2 of 2 answered).

2026-11-06: 1 of 1 checks answered
2026-11-07: 1 of 1 checks answered

30-day availability

Aggregated from read-only remote checks; this is not a tool-execution or security guarantee.

Successful checks: 1 of 1 (100%)
Average latency: 496 ms
95th percentile latency: 496 ms

Maintenance history

Public check and source-refresh events. Editorial and ownership records remain private.

Nov 6, 2026, 3:22 PM UTC · probe · passed · 496 ms · metadata changed · 1 tool(s)

Connect from your client

Each snippet registers the same endpoint; pick the client you use.

This server is operated by the named third party. Listing and endpoint checks are not an endorsement or a security audit. Review the operator's permissions, terms, and data handling before connecting. How listings are checked and corrected.

Claude Code

Run this once in a terminal. It registers the remote server. Use --scope project to write .mcp.json for the repo.

Claude Code configuration

claude mcp add --transport http syftly-mcp https://syftly.vercel.app/api/mcp

Codex

Run this once. Codex stores MCP servers in ~/.codex/config.toml (or .codex/config.toml in a trusted project).

Codex configuration

codex mcp add syftly-mcp --url https://syftly.vercel.app/api/mcp

Cursor

Merge this server into .cursor/mcp.json in your project, then enable syftly-mcp in your MCP settings.

Cursor configuration

{
  "mcpServers": {
    "syftly-mcp": {
      "url": "https://syftly.vercel.app/api/mcp"
    }
  }
}

Cline

Merge this into cline_mcp_settings.json (Cline → MCP Servers → Configure).

Cline configuration

{
  "mcpServers": {
    "syftly-mcp": {
      "type": "streamableHttp",
      "url": "https://syftly.vercel.app/api/mcp",
      "disabled": false
    }
  }
}

VS Code

Merge this server into .vscode/mcp.json. Use MCP: List Servers to start syftly-mcp and inspect its tools.

VS Code configuration

{
  "servers": {
    "syftly-mcp": {
      "type": "http",
      "url": "https://syftly.vercel.app/api/mcp"
    }
  }
}

Gemini CLI

Run this once. It writes .gemini/settings.json in the current project; pass --scope user for ~/.gemini/settings.json.

Gemini CLI configuration

gemini mcp add --transport http syftly-mcp https://syftly.vercel.app/api/mcp

Claude Desktop

Open Claude Settings > Connectors > Add custom connector, enter this server URL and enable it for the conversation. This is a manual remote-connector handoff, not a desktop JSON configuration.

Claude Desktop configuration

https://syftly.vercel.app/api/mcp

Windsurf

Merge this into ~/.codeium/windsurf/mcp_config.json, then refresh MCP in Cascade. Windsurf remote servers use serverUrl, not url.

Windsurf configuration

{
  "mcpServers": {
    "syftly-mcp": {
      "serverUrl": "https://syftly.vercel.app/api/mcp"
    }
  }
}

Let your agent find its tools

Find servers like Syftly MCP from your agent. Add Koskamo once to search the index for endpoints, access requirements, and client setup. Public discovery needs no key.

Connect Koskamo Explore workflows

See all 5,159