# Groq MCP connector Connecting Groq to Elaichi lets Claude, ChatGPT, Cursor, or any MCP client transcribe and translate audio, generate speech, run chat completions, and manage batches, files, and fine tunings, with every request logged under the person who made it. Source: https://elaichi.ai/connectors/groq/ ## Facts | | | | --- | --- | | Application | Groq | | Category | Conversational Intelligence | | AI tools | 21 | | Authentication | Connects with an API key | | Needs your own OAuth app | No | | MCP endpoint | https://api.elaichi.ai/mcp | | Works with | Claude, ChatGPT, Cursor, any MCP client, and the Elaichi Agent | | Tools advertised by name | Yes | ## What you can ask once Groq is connected - Summarize this week's Groq transcriptions by team. - Which Groq batches are still running right now? - List the Groq models available for audio transcription. ## Connect Groq in Elaichi This happens once for the organization, before any client is involved. 1. Open Connections, choose Add connection, and pick Groq. 2. Optionally set Share with, then press Connect. 3. Paste a Groq API key. One person generates a token in Groq and pastes it once. Everyone else works through Share with, and never sees it. Credentials are vaulted and nobody, including the AI, reads them back. The connection becomes a toolbox immediately, so you can curate which Groq tools are exposed, rename them, or freeze arguments before anyone points a client at it. ## Connect Groq to Claude Endpoint: https://api.elaichi.ai/mcp 1. Open Customize, then Connectors. 2. Press Add. 3. Name it, paste the MCP server URL, then Continue. 4. Sign in and approve. On Team and Enterprise, an Owner adds it once. Everyone else turns it on for themselves. ## Connect Groq to ChatGPT Endpoint: https://api.elaichi.ai/mcp 1. Open Plugins, then press the + button. 2. Name it and paste the endpoint into Server URL. 3. Leave Authentication on OAuth, then tick the risk acknowledgement. 4. Press Create, then sign in and approve. Works on the web today. The plugin directory lives at chatgpt.com/plugins. ## Connect Groq to Cursor Endpoint: https://api.elaichi.ai/mcp 1. Open `~/.cursor/mcp.json`. 2. Add the endpoint under `mcpServers`. 3. Reload Cursor, then sign in and approve. Set up per machine, so repeat it on each computer you work from. ## Connect Groq to any MCP client Endpoint: https://api.elaichi.ai/mcp 1. Add the endpoint as a remote MCP server. 2. Sign in and approve. The Elaichi Agent already has these tools, with nothing to set up. ## What the consent screen decides Only Read is granted by default, which is not enough to call a Groq tool. Over MCP there is no trusted place to confirm a write in the moment, so the consent screen is the standing approval rather than a formality. Grant Read and Run tools. Think hard before granting Delete, which reaches into connected apps and cannot be undone. ## What teams do with Groq through Elaichi ### Turn recorded calls into searchable text Support. Send a customer call recording to Groq for transcription and get back a text version the desk can search, quote, and attach to the ticket. ### Translate a prospect call into English Sales. Upload a demo recording in another language and have Groq translate it into English text so the whole team can read what was promised. ### Give a blog post a spoken version Marketing. Hand Groq the final copy and get an audio file back, ready to publish alongside the article or drop into a video. ### Compare available models before a launch Product. Ask which Groq models are available right now and pull the details on one before deciding what a new feature should run on. ### Run a large job overnight as a batch Data. Upload a file of prompts, start a Groq batch, check on its progress in the morning, and cancel it if the inputs were wrong. ### Keep files and fine tunings tidy Operations. List the files and fine tunings sitting in the Groq account, download the ones still needed, and delete the rest so nothing lingers. ## Frequently asked questions ### How do I connect Groq to Claude? In Elaichi, choose Groq and paste in the API key from your Groq account. There is no OAuth application to register and no client ID or secret to generate. Then in Claude, go to Customize, then Connectors, then Add, and paste the endpoint https://api.elaichi.ai/mcp. Sign in with your Elaichi account and Groq is ready to use in Claude. ### Does Groq work with ChatGPT and Cursor as well as Claude? Yes. Once Groq is connected in Elaichi, the same endpoint, https://api.elaichi.ai/mcp, works in Claude, ChatGPT, Cursor, any other MCP client, and the Elaichi Agent. You connect Groq once and every client you use picks it up. ### What can an AI agent actually do with Groq? An agent connected to Groq can transcribe an audio recording, translate speech into English text, turn written text into spoken audio, and run chat completions or responses on a chosen model. It can also list the models available to you, upload and download files, start, check, and cancel batch jobs, and create or remove fine tunings. ### Does connecting Groq give the AI everything in my Groq account? An agent can only do what the Groq API key it was connected with is allowed to do, and Elaichi records every request under the person who made it. Elaichi can narrow that access further, for example by blocking deletions or fine tuning, but it can never widen it beyond what the Groq key already permits. ### Can my team share one Groq connection? Yes. One person connects Groq with the API key, shares the connection with a team in Elaichi, and nobody else ever sees or handles that key. Each teammate still signs in to Elaichi as themselves, so the audit log names the individual behind every Groq transcription, batch, or file change. ### Can I stop an agent from deleting or changing things in Groq? Yes. Restrictions in Elaichi work per action, so you can allow transcription and chat completions in Groq while blocking file deletion, fine tuning removal, or batch cancellation. A blocked action is never shown to Claude, ChatGPT, Cursor, or any other client, so no prompt can reach it. ### What happens to a Groq connection when someone leaves? Offboarding a person in Elaichi ends their access to Groq immediately, across every client they used. If they shared a Groq connection with a team, it keeps working for everyone else. Disconnecting Groq once in Elaichi removes it from Claude, ChatGPT, Cursor, and every other connected client at the same time. ## All 21 Groq tools Every tool below is callable through https://api.elaichi.ai/mcp once Groq is connected, subject to the toolbox it is in and the restrictions on the caller. - **Create a Groq audio transcription** (Create). Transcribe an audio file into text using groq's audio transcription API. Returns: text (the full transcription), and when response_format is verbose_json also task, language, duration, segments, words, and x_groq metadata. Required: file, model. - **Create a Groq audio translation** (Create). Create an audio translation in groq by uploading an audio file, converting spoken content into English text. Returns: text. Required: file, model. - **Create a Groq transcription** (Create). Create an audio transcription in groq by uploading an audio file or supplying a URL to convert speech to text. Returns: text, segments, words. Required: model. Either file or url must also be provided. - **Create a Groq chat completion** (Create). Create a chat completion in Groq using a specified model and message array, with optional assistant-prefill and stop-sequence control. Returns: id, object, model, choices (containing the assistant message, finish_reason, and logprobs), usage (prompt_tokens, completion_tokens, total_tokens, prompt_time, completion_time, total_time), and x_groq. Required: messages, model. - **Create a Groq response** (Create). Create a model response using Groq's OpenAI-compatible Responses API. Accepts a model identifier and input (text or message array) along with optional sampling, tool, and reasoning parameters. Returns: id, object, model, output, usage, status, created_at, and metadata describing the completed response. - **Create a Groq audio speech** (Create). Create synthesized audio speech from text using groq's OpenAI-compatible text-to-speech API. Returns binary audio data in the requested format (not a JSON body). Required: model, input, voice. - **List all Groq models** (List). List all models available on groq. Returns a collection of model objects, each including id, object, created, and owned_by. - **Get single Groq model by ID** (Get). Get a single groq model by id. Returns the model object including id, object, created, and owned_by. Required: id. - **List all Groq batches** (List). List all batches in groq. Returns a collection of batch objects each including id, object, status, endpoint, input_file_id, completion_window, created_at, request_counts, and associated file identifiers. - **Create a Groq batch** (Create). Create a new batch in groq to process a group of requests asynchronously. Returns the created batch object including id, status, endpoint, input_file_id, completion_window, created_at, and request_counts. Required: input_file_id, endpoint, completion_window. - **Get single Groq batch by ID** (Get). Get a single groq batch by id. Returns the full batch object including id, status, endpoint, input_file_id, completion_window, created_at, request_counts, output_file_id, and all lifecycle timestamps. Required: id. - **Groq batches cancel** (Cancel). Cancel an in-progress groq batch by id. Returns the updated batch object reflecting the cancellation, including id, status (cancelling or cancelled), endpoint, input_file_id, completion_window, created_at, and request_counts. Required: id. - **List all Groq files** (List). List all files uploaded to groq. Returns: id, object, bytes, created_at, filename, purpose for each file in the account. - **Create a Groq file** (Create). Upload a file to groq for use with other API endpoints. Returns: id, object, bytes, created_at, filename, purpose. Required: file, purpose. - **Get single Groq file by ID** (Get). Get metadata for a specific groq file by id. Returns: id, object, bytes, created_at, filename, purpose. Required: id. - **Delete a Groq file by ID** (Delete). Delete a groq file by id. Returns: id, object, deleted. Required: id. - **Groq files download** (Download). Download the raw content of a groq file by id. Returns the file's binary or text payload; the response shape depends on the file's content type. Required: id. - **List all Groq fine tunings** (List). List fine-tuning jobs in groq. Returns: id. - **Create a Groq fine tuning** (Create). Create a new fine-tuning job in groq. Returns: id. - **Get single Groq fine tuning by ID** (Get). Get a single groq fine-tuning job by id. Returns: id. Required: id. - **Delete a Groq fine tuning by ID** (Delete). Delete a groq fine-tuning job by id. Returns an empty 204 response on success. Required: id.