Skip to main content
Glama

kurdish-tts-stt

by com.kurdishtts
OverviewDiscussionsAdmin

Kurdish speech to text

transcribe_audio

Transcribe Kurdish audio (Sorani or Kurmanji) to text. Requires an STT API key; usage is metered per audio minute against your plan. IMPORTANT: dialect selects the decoder and nothing detects it for you — transcribing Sorani audio as Kurmanji returns fluent, confident, WRONG text with no error. Pass dialect "auto" when you are not certain, and pick the coherent transcript from the two it returns. Max 3MB of decoded audio over MCP (~90s of 16kHz WAV, but ~25 minutes of 64kbps MP3 — send compressed audio to fit more); for larger files call POST https://www.kurdishtts.com/api/stt-proxy directly (multipart).

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
dialectYesWhich decoder to run. "sorani" or "kurmanji" when you know the dialect. "auto" transcribes with BOTH and returns both transcripts so you can choose — it bills the audio twice, so prefer a known dialect when you have one.
filenameNoOriginal filename, used to infer the format. Default audio.wav.
mime_typeNoAudio MIME type. Default audio/wav.
audio_base64YesBase64-encoded audio file (WAV, MP3, FLAC, OGG or M4A). Max 3MB decoded — compressed formats fit far more speech in that budget than WAV does.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed3 schema fields changed
    • changedInput schema / properties / audio_base64 / description
      Previous value: -"Base64-encoded audio file (WAV, MP3, FLAC, OGG or M4A). Max 3MB decoded."New value: +"Base64-encoded audio file (WAV, MP3, FLAC, OGG or M4A). Max 3MB decoded — compressed formats fit far more speech in that budget than WAV does."
    • changedInput schema / properties / dialect / description
      Previous value: -"Dialect spoken in the audio."New value: +"Which decoder to run. \"sorani\" or \"kurmanji\" when you know the dialect. \"auto\" transcribes with BOTH and returns both transcripts so you can choose — it bills the audio twice, so prefer a known dialect when you have one."
    • changedInput schema / properties / dialect / enum
      Previous value: -[
-  "sorani",
-  "kurmanji"
-]New value: +[
+  "sorani",
+  "kurmanji",
+  "auto"
+]
  2. Changed5 schema fields changed
    • addedInput schema / additionalProperties
      Added value: +false
    • changedInput schema / properties / audio_base64 / description
      Previous value: -"Base64-encoded audio file (WAV, MP3, FLAC, OGG or M4A)."New value: +"Base64-encoded audio file (WAV, MP3, FLAC, OGG or M4A). Max 3MB decoded."
    • addedInput schema / properties / audio_base64 / maxLength
      Added value: +4194308
    • removedInput schema / properties / context
      Removed value: -{
-  "description": "Explain why you are calling this tool and how it fits into the user's overall goal. This parameter is used for analytics and user intent tracking. YOU MUST provide 15-25 words (count carefully). NEVER use first person ('I', 'we', 'you') - maintain third-person perspective. NEVER include sensitive information such as credentials, passwords, or personal data. Example (20 words): \"Searching across the organization's repositories to find all open issues related to performance complaints and latency issues for team prioritization.\"",
-  "type": "string"
-}
    • changedInput schema / required
      Previous value: -[
-  "audio_base64",
-  "dialect",
-  "context"
-]New value: +[
+  "audio_base64",
+  "dialect"
+]
  3. First observed

TDQS

A4.5/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Goes well beyond the annotations (which only declare non-read-only, non-destructive, non-idempotent) by disclosing the API-key requirement, per-minute metering, the double billing of "auto", and the silent wrong-output failure mode when the dialect is misidentified. This is exactly the context an agent cannot get from structured fields.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Front-loads the operation, then the hard constraint (dialect selection), then the size limit and escape hatch. Dense but every clause earns its place; the parenthetical bitrate arithmetic is the only slightly bulky part.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

No output schema exists, and the description compensates by stating what "auto" returns (two transcripts) and warning about the silent-failure mode. It stops short of describing the single-dialect response shape or error behavior, a minor gap for a well-annotated tool.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema coverage is 100%, so the baseline is 3, but the description adds real meaning: it explains the consequence of the dialect choice (wrong decoder produces confident nonsense) and the practical size math for WAV versus MP3. It does not add anything about filename or mime_type, which the schema already handles.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

States a specific verb and resource ('Transcribe Kurdish audio ... to text') and narrows scope to two dialects, which cleanly separates it from synthesize_speech and the streaming sibling. An agent can identify the operation immediately.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Gives explicit conditions for choosing a known dialect versus "auto" and tells the agent to bypass MCP for files over the size limit, including the endpoint and encoding. It never names the sibling start_streaming_transcription as the alternative for long audio, so routing within the toolset is left to inference.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

All kurdish-tts-stt toolsInspect
Try in Browser
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

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources