# Groq > Groq through One - Skill: `withoneai/groq` (Agent Skill) - Install (CLI): `npx skillmds@latest add withoneai/groq` - Raw SKILL.md: https://api.skillmd.com/api/skills/withoneai/groq/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - Author: withoneai (https://skillmd.com/u/withoneai) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/withoneai/groq --- # Groq through One Groq is an AI inference platform that provides high-performance infrastructure and APIs for running large language models with low latency, allowing developers and enterprises to build and scale real-time AI applications, assistants, and model-powered workflows. One exposes Groq through four MCP tools. The table below carries real action ids from One's knowledge base, so for a common operation you can skip search and go straight to reading the action's parameters. ## How to run an action 1. Find the action in the table below, or call `search_one_platform_actions` with platform `groq` if it is not listed. 2. Call `get_one_action_knowledge` with the action id. Do this every time, including for actions in this table. The table gives you the id, not the parameters. 3. Call `execute_one_action` with parameters copied from that knowledge. Never guess a parameter name, a body field, or an enum value. The knowledge has the real schema, and a guessed field is either a 400 or a silent write of the wrong thing. ## Before you start Call `list_one_integrations` once and confirm Groq is connected. If it is missing, the user has not connected it: say so and point them at https://app.withone.ai rather than reaching for raw HTTP. Each connection carries an `access` field. If it reports `{"policy": "methods", "methods": ["GET"]}` the agent is read-only here, so plan a read-only answer instead of attempting a write that will be refused. ## Before a write Creates, updates, deletes and sends land on a real Groq account and cannot be recalled. State the action and the specific target in one line before the first write in a task, and let the user stop you. Reads need no confirmation. ## Actions ### Files | Action | Method | Path | Action id | |---|---|---|---| | Download File Content | GET | `/openai/v1/files/{{fileId}}/content` | `conn_mod_def::GLp2vyw6RKI::AeC9mS8OQGWH-VU3x11x5Q` | | List Files | GET | `/openai/v1/files` | `conn_mod_def::GLp2vzWwkJo::wG3aHjSmQ4KL3bGPQtbi5Q` | | Retrieve a File | GET | `/openai/v1/files/{{fileId}}` | `conn_mod_def::GLp2vysO-mE::ZW99-o2dToKVrr7D1cQ9CQ` | | Delete a File | DELETE | `/openai/v1/files/{{fileId}}` | `conn_mod_def::GLp2vonGuaE::eBQntJ1CTiCry5wfP8VAfQ` | | Upload a File | POST | `/openai/v1/files` | `conn_mod_def::GLp2vzVLoss::FgOvli6QRhyFpAxZtaNSyg` | ### Batches | Action | Method | Path | Action id | |---|---|---|---| | List Batches Using OpenAI v1 | GET | `/openai/v1/batches` | `conn_mod_def::GLp2vtjVimM::qXGLPLQ2S6StbPTfU8LItw` | | Retrieve Batch | GET | `/openai/v1/batches/{{batchId}}` | `conn_mod_def::GLp2vpVzdoA::qYLV2djnQJqecYhTUzDGUA` | | Cancel a Batch | POST | `/openai/v1/batches/{{batchId}}/cancel` | `conn_mod_def::GLp2vhI7aM0::JjoW9Ka0SWaeXM-wTNQtQQ` | | Create Batch Using OpenAI v1 | POST | `/openai/v1/batches` | `conn_mod_def::GLp2vhfKr78::wfJXO35zRAKAuF3_C-1xBg` | ### FineTunings | Action | Method | Path | Action id | |---|---|---|---| | List Fine Tunings | GET | `/v1/fine_tunings` | `conn_mod_def::GLp2v404OXg::C7rU9eKjSkeA-FRkJNWDaA` | | Retrieve a Fine Tuning by ID | GET | `/v1/fine_tunings/{{id}}` | `conn_mod_def::GLp2v7HlWeA::zK0b2rv3ThiQbC6kOe0kig` | | Create Fine Tunings | POST | `/v1/fine_tunings` | `conn_mod_def::GLp2vzQyc-4::t_Zam-05QvC0yY866OnXqg` | | Delete a Fine Tuning | DELETE | `/v1/fine_tunings/{{id}}` | `conn_mod_def::GLp2v4sFuvc::Dy7mlyBIQcirNKmIRh4cWA` | ### Models | Action | Method | Path | Action id | |---|---|---|---| | Get a Model | GET | `/openai/v1/models/{{model}}` | `conn_mod_def::GLp2v48V_kw::h3PJ69bTTrynBXIBfEwgEQ` | | List Available Models | GET | `/openai/v1/models` | `conn_mod_def::GLp2wDXvMmA::xDxh2JgNRoSUlrt1wCY5qQ` | | Delete Model | DELETE | `/openai/v1/models/{{model}}` | `conn_mod_def::GLp2v6g-1NQ::yY3-qMZBQVaS3AAUq-b2Kw` | ### AudioSpeech | Action | Method | Path | Action id | |---|---|---|---| | Speech Using OpenAI v1 Audio | POST | `/openai/v1/audio/speech` | `conn_mod_def::GLp2vg5inAo::ccaD_GyoT_2dbftkvUCYyQ` | ### AudioTranscriptions | Action | Method | Path | Action id | |---|---|---|---| | Create Audio Transcription | POST | `/openai/v1/audio/transcriptions` | `conn_mod_def::GLp2vhQ5i_0::sAKnecXBTCaHf0C1IVBitA` | ### AudioTranslations | Action | Method | Path | Action id | |---|---|---|---| | Translate Audio into English | POST | `/openai/v1/audio/translations` | `conn_mod_def::GLp2virvqU0::uZdFDo5BQAurHyMC5KhUvQ` | ### ChatCompletions | Action | Method | Path | Action id | |---|---|---|---| | Create Chat Completion | POST | `/openai/v1/chat/completions` | `conn_mod_def::GLp2vos0PN8::mag-A6R4Q_Gz22i4fmRQ7A` | ### Embeddings | Action | Method | Path | Action id | |---|---|---|---| | Create Embedding Using OpenAI v1 | POST | `/openai/v1/embeddings` | `conn_mod_def::GLp2vpVliZw::VqIIGRVMSyCbkZZT3Zi4GA` | ### Responses | Action | Method | Path | Action id | |---|---|---|---| | Create Response | POST | `/openai/v1/responses` | `conn_mod_def::GLp2wAkepFY::kEiCA00tR1yr1kruS1sJ4Q` | ### Reranking | Action | Method | Path | Action id | |---|---|---|---| | Reranking Using OpenAI v1 | POST | `/openai/v1/reranking` | `conn_mod_def::GLp2wBptS7M::VQTtMFKqQkKpWvY5SSp6ug` | ## When a call fails The error comes from Groq, not from One. A 400 or 422 means your parameters do not match the schema, so re-read the knowledge and fix the field. A 401 or 403 means the connection needs re-authorizing, which no retry will fix. A 404 means the id is not on this account. A 429 means slow down. Never retry a write more than once: the first attempt may have landed. Full catalog: https://www.withone.ai/knowledge/groq Integration knowledge by One (withone.ai), licensed under [One-Knowledge-1.0](https://www.withone.ai/licenses/knowledge). Attribution must be preserved in derivative works.