Convert text to spoken audio. Returns a status with byte size.
Args:
text: the text to speak.
voice: ⚠️ 当前默认模型 (qwen3-tts-flash) **忽略 OpenAI 音色名** ——
alloy / echo / fable / onyx / nova / shimmer 六个会被统一改写成同一个音色,
传哪个都得到逐字节相同的音频。要真正换音色请传上游音色名 (如 Cherry / Ethan)。
| Name | Required | Description | Default |
|---|---|---|---|
| text | Yes | ||
| voice | No | alloy |
| Name | Required | Description | Default |
|---|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries full behavioral burden. It goes beyond a generic statement by exposing a critical quirk: the default model ignores OpenAI voice names and normalizes all six to the same audio, while also explaining how to achieve a real voice change via upstream names. This is exactly the kind of non-obvious behavior an agent needs to know before invoking the tool.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is compact and front-loaded with the core purpose. The parameter notes are structured and the extended voice warning is dense but necessary. It is slightly verbose due to the inline warning and code-like formatting, but every part earns its place.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a two-parameter tool with an output schema present, the description covers the core invocation details: what to pass, what unusual voice behavior to expect, and what kind of result is returned. It does not mention audio format, delivery mechanism, or text length constraints, but the presence of an output schema lowers the burden for return-value details.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 0%, so the description must explain the parameters, and it does: 'text' is defined as the text to speak, and 'voice' receives a detailed warning about name normalization and examples of valid upstream names. It adds substantial meaning beyond the bare schema, although it does not enumerate all possible voice values or text length limits.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description opens with a precise verb and resource, 'Convert text to spoken audio', which fully defines the tool's function. It also distinguishes it from its siblings (chat, image/video generation, status, list_models) without needing to name them. The added detail about returning a status with byte size further clarifies observable behavior.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The main sentence implies when to use this tool: whenever text needs to be turned into speech. It does not, however, explicitly state when not to use it or compare it to alternatives such as generate_video or chat. The usage context is inferable but not spelled out.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Add one secure layer between your agents and this server.
Each tool maps to a distinct modality or lifecycle step: chat, image generation, video generation, video status polling, model listing, and speech synthesis. The async video pair (generate_video/get_video_status) is clearly separated, and no two tools plausibly do the same thing.
Most tools follow a clear verb_noun pattern: generate_image, generate_video, get_video_status, list_models. The deviations are minor: 'chat' is a bare verb and 'text_to_speech' uses a noun_phrase style instead of generate_speech or synthesize_speech, but the overall naming is still predictable.
Six tools is well-scoped for a multimodal AI gateway covering chat, image, video, speech, and model discovery. Each tool earns its place and the count is comfortably within the ideal range.
The core workflows are covered: chat has a direct call, video generation has an async create-and-poll lifecycle, image and speech have generation entry points, and list_models aids discovery. Minor gaps exist—no image-to-video, speech status polling, or model selection for image/video/speech—but agents can accomplish the primary stated purposes.