openai-mcp
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| OPENAI_API_KEY | Yes | Your OpenAI API key. Required; a missing or invalid key is reported as an error on the first tool call. | |
| OPENAI_MCP_TEXT_MODELS | No | Enables generate_text and count_tokens. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_IMAGE_MODELS | No | Enables generate_image and edit_image. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_SPEECH_MODELS | No | Enables generate_speech. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_EMBEDDING_MODELS | No | Enables create_embeddings. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_MODERATION_MODELS | No | Enables create_moderation. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_TRANSLATION_MODELS | No | Enables translate_audio. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. | |
| OPENAI_MCP_TRANSCRIPTION_MODELS | No | Enables transcribe_audio. Comma-separated model IDs, or * to allow any model. Omit to leave the tool disabled. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| generate_textA | Send a prompt to an OpenAI model and return its text reply. Pass images to ask a vision-capable model about them |
| count_tokensA | Count the input tokens a generate_text request would use, without running the model |
| generate_imageA | Generate an image from a text prompt using an OpenAI image model. By default returns the local file path(s) the image was saved to rather than inline image data; set returnAs to 'base64' for inline data instead |
| edit_imageA | Edit or combine existing images using a text prompt and an OpenAI image model. By default returns the local file path(s) the result was saved to rather than inline image data; set returnAs to 'base64' for inline data instead |
| create_embeddingsC | Get vector embeddings for one or more pieces of text |
| list_modelsA | List available OpenAI models. Returns configured models unless a tool allows any model |
| generate_speechA | Convert text to spoken audio using an OpenAI text-to-speech model. By default returns the local file path the audio was saved to rather than inline data; set returnAs to 'base64' for inline audio instead |
| transcribe_audioB | Transcribe spoken audio to text using an OpenAI speech-to-text model |
| translate_audioA | Translate spoken audio in any supported language into English text using an OpenAI speech-to-text model. Output may occasionally come back untranslated, in its original language |
| create_moderationA | Check text for potentially harmful content using an OpenAI moderation model. Returns, for each input, whether it was flagged plus per-category flags and scores |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
Most tools target clearly distinct OpenAI capabilities: text generation, token counting, embeddings, moderation, image generation/editing, model listing, and speech generation/transcription/translation. The only mild overlap is transcribe_audio versus translate_audio, since both convert speech to text, though the descriptions make the translation-to-English distinction reasonably clear.
All names are snake_case and follow a verb_noun structure, which is highly predictable. There is a minor inconsistency in verb choice between create_embeddings/create_moderation and generate_text/generate_image/generate_speech, but the pattern remains readable.
Ten tools is a well-scoped set for an OpenAI API wrapper, covering major modalities without obvious filler. Each tool maps to a recognizable OpenAI capability and earns its place.
The surface covers text generation, vision input, token counting, embeddings, moderation, image generation/editing, model listing, TTS, transcription, and translation. Minor gaps exist around other OpenAI API areas such as fine-tuning, files, or assistants, but core model-inference workflows are well covered.