io.github.AceDataCloud/mcp-glm

Official

Overview Schema Source Related Servers Discussions Admin

glm_chat_completions

Sends messages to a selected Zhipu GLM model and returns the generated chat completion, enabling chat, Chinese language understanding, and reasoning tasks.

Instructions

Create a GLM chat completion using the AceDataCloud GLM API.

Sends messages to the specified GLM model and returns the generated response.
Supports all GLM models: glm-5.3, glm-5.2, glm-5, glm-5-turbo, glm-5.1, glm-4.7, glm-4.6, glm-3-turbo.

Use this when:
- You need to chat with a Zhipu GLM model
- You need Chinese language understanding or generation
- You want to use GLM's reasoning capabilities

Returns:
    JSON response containing the chat completion result.

Input Schema

Name	Required	Description	Default
`n`	No	How many chat completion choices to generate. Default 1.
`seed`	No	Random seed for deterministic sampling.
`stop`	No	Stop sequences where the API will stop generating tokens.
`user`	No	End-user identifier for abuse monitoring.
`audio`	No	Parameters for audio output.
`model`	No	The GLM model to use. Options: glm-5.3, glm-5.2, glm-5, glm-5-turbo, glm-5.1, glm-4.7, glm-4.6, glm-3-turbo. Default is glm-5.2.	glm-5.2
`store`	No	Whether to store the output of this request. Default False.
`tools`	No	List of tools the model may call.
`top_p`	No	Nucleus sampling probability mass. Default 1.
`stream`	No	This MCP tool returns a complete JSON response. Omit or set false; stream=true is rejected before sending a request.
`logprobs`	No	Whether to return log probabilities of output tokens. Default False.
`messages`	Yes	Conversation messages. Each message is a dict with 'role' and 'content' keys. Required.
`metadata`	No	Key-value pairs for storing additional information.
`logit_bias`	No	Modify the likelihood of specified tokens appearing in the completion.
`max_tokens`	No	Maximum number of tokens to generate.
`modalities`	No	Output types to generate (e.g. ['text', 'audio']).
`prediction`	No	Static predicted output content for latency reduction.
`temperature`	No	Sampling temperature between 0 and 2. Higher = more random. Default 1.
`tool_choice`	No	Controls which tool is called. Can be 'none', 'auto', or a dict.
`service_tier`	No	Service tier: auto, default, flex, scale, priority. Default auto.
`top_logprobs`	No	Number of most likely tokens to return at each token position.
`stream_options`	No	Options for streaming response.
`response_format`	No	Response format specification (e.g. {"type": "json_object"}).
`presence_penalty`	No	Presence penalty between -2.0 and 2.0. Positive values increase topic variety. Default 0.
`reasoning_effort`	No	Reasoning effort level: minimal, low, medium, high. Default medium.
`frequency_penalty`	No	Frequency penalty between -2.0 and 2.0. Positive values decrease repetition. Default 0.
`web_search_options`	No	Options for web search tool.
`parallel_tool_calls`	No	Enable parallel function calling. Default True.
`max_completion_tokens`	No	Upper bound for tokens that can be generated for a completion.

Output Schema

Name	Required	Description	Default
`result`	Yes

Schema Changelog

Changes observed during successful MCP inspections.

Changed1 schema field changedv0.1.8Nov 4, 2026

changedInput schema / properties / stream / description

Previous value: -"Whether to stream partial message deltas. Default False."New value: +"This MCP tool returns a complete JSON response. Omit or set false; stream=true is rejected before sending a request."

Changed2 schema fields changedv0.1.2Aug 29, 2026

changedInput schema / properties / model / description

Previous value: -"The GLM model to use. Options: glm-5.2, glm-5, glm-5-turbo, glm-5.1, glm-4.7, glm-4.6, glm-3-turbo. Default is glm-5.2."New value: +"The GLM model to use. Options: glm-5.3, glm-5.2, glm-5, glm-5-turbo, glm-5.1, glm-4.7, glm-4.6, glm-3-turbo. Default is glm-5.2."

changedInput schema / properties / model / enum

Previous value: -[
-  "glm-5.2",
-  "glm-5",
-  "glm-5-turbo",
-  "glm-5.1",
-  "glm-4.7",
-  "glm-4.6",
-  "glm-3-turbo"
-]New value: +[
+  "glm-5.3",
+  "glm-5.2",
+  "glm-5",
+  "glm-5-turbo",
+  "glm-5.1",
+  "glm-4.7",
+  "glm-4.6",
+  "glm-3-turbo"
+]

First observedv0.1.0Aug 7, 2026

TDQS

B3.4/5.0

Behavior2/5

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

No annotations are supplied, so the description carries the full behavioral burden, and it does not disclose side effects, billing/quota consumption, persistence semantics of store=true, statelessness between calls, or any auth/rate-limit constraints. It only restates the basic request-response flow ('sends messages ... returns the generated response'), which is already implied by the name and title. It does not contradict anything, but it adds little real behavioral context.

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-loaded with purpose, then a scannable 'Use this when' list and a Returns section, so an agent can extract the essentials quickly. The model list is repeated verbatim from the schema enum, which is the one clearly redundant sentence, and the 'Returns:' section is nearly empty filler.

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

Completeness3/5

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

An output schema exists, so omitting detailed return-value documentation is acceptable, and the description covers the basic what and when. For a 29-parameter tool with no annotations, though, it leaves out model-selection tradeoffs, the fact that streaming is rejected, cost/persistence implications of store, and error behavior, leaving the agent to infer these from the schema alone.

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

Parameters3/5

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

Schema description coverage is 100%, so all 29 parameters are already documented in the schema, and the baseline of 3 applies. The description's only parameter-adjacent content is re-listing the model enum values, which duplicates structured data rather than adding selection guidance or interaction notes (e.g., max_tokens vs max_completion_tokens).

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

Purpose4/5

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

States a specific verb and resource ('Create a GLM chat completion'), names the upstream API, and explains that it sends messages to a model and returns the generated response. The purpose is unambiguous and clearly distinct from glm_list_models and glm_get_usage_guide, though it never explicitly contrasts itself with those siblings.

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?

The 'Use this when' block gives concrete triggering contexts (chatting with a Zhipu GLM model, Chinese language tasks, reasoning workloads), which is more than most descriptions offer. However, there is no when-not guidance, no alternative tool named, and no advice on choosing among the eight listed models (e.g., turbo for speed vs. glm-5.3 for quality).

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

Deploy Server