Kaggle MCP Server

Endpoint

Hosted endpoint · transport

Connect

This is a remote MCP server — point an MCP client at the endpoint URL. Clients with native remote support (Claude, Cursor, VS Code, …):

For stdio-only clients (older Claude Desktop, etc.), bridge with mcp-remote:

If the server requires authentication, add the provider's token/header (e.g. an X-API-Key) per its docs. A quick reachability check:

Tools

• get_competition_data_files_summary — Get a summary of the data files for a competition, including number and type of files as well as column metadata.
• update_benchmark_arena_combinations — Opt task and candidate combinations into or out of a benchmark version's arena.
• get_competition_leaderboard — Get the leaderboard for a competition.
• update_dataset_metadata — Update the metadata for a dataset.
• download_dataset — Download a dataset.
• search_content — Search across all Kaggle content types (competitions, datasets, notebooks, models, discussions, users, benchmarks).
• download_competition_leaderboard — Download the leaderboard for a competition.
• list_forum_topics — List and search discussion topics. Omit forum_id to search across all forums.
• get_model_variation — Get metadata about a model variation.
• get_dataset_files_summary — Get a summary of the data files in a dataset, including number and type of files as well as column metadata.
• list_model_variation_versions — List model variation versions.
• get_benchmark_leaderboard — Get the leaderboard for a benchmark.
• list_benchmark_arena_combinations — List the task and candidate combinations opted into a benchmark version's arena.
• create_benchmark_arena_settings — Create a benchmark version's arena.
• list_hackathon_write_ups — List hackathon write-up submissions for a competition. Supports filtering by track and winner status.
• download_model_variation_version — Get a url where the files for a model variation version can be downloaded.
• get_notebook_info — Get information about a notebook.
• list_forums — List all top-level discussion forums on Kaggle.
• create_code_competition_submission — Submit a kernel to a competition.
• list_hackathon_tracks — List the tracks for a hackathon competition.
• get_accelerator_quota — Get the current user's weekly GPU and TPU quota usage.
• list_models — List models.
• create_model — Create a new model.
• get_episode_replay — Download the replay for a simulation episode.
• get_dataset_status — Get the processing status of a dataset, indicating whether the dataset is still processing, completed, or failed.
• create_benchmark_task_from_prompt — Create a new benchmark task from a prompt.
• get_notebook_session_status — Get the status of a notebook session, indicating whether it is running, completed, or failed..
• download_notebook_output_zip — Download the zipped output of a notebook.
• update_model — Update a model.
• download_competition_submission — Download the file for one of your competition submissions.
• get_forum_topic — Get a discussion topic by ID, optionally including its comments.
• get_user_profile — Get a user's public profile.
• get_benchmark_arena_settings — Get a benchmark version's arena settings.
• get_model — Get a metadata about a model.
• get_writeup_by_topic — Get a writeup by its linked forum topic ID. Use this when you found a writeup via forum discussions.
• download_competition_data_file — Download a single data file for a competition.
• get_dataset_info — Get information about a dataset.
• get_episode_agent_logs — Download agent logs for a simulation episode.
• list_model_variation_version_files — List the files of a model variation version.
• list_competition_pages — List the content pages for a competition (overview, rules, evaluation, data description, etc.).
• get_competition_submission — Get metadata about a competition submission.
• cancel_notebook_session — Cancel a notebook session.
• authorize — This tool can be used to authorize with Kaggle but it is only compatible with certain clients. See kaggle.com/docs/mcp for more information
• get_hackathon_overview — Get all the overview page content for a hackathon competition.
• list_competition_data_files — List the data files for a competition.
• update_benchmark_arena_settings — Update a benchmark version's arena settings.
• search_competition_submissions — Search for competition submissions.
• get_dataset_metadata — Get metadata about a dataset.
• list_team_public_submissions — List the public-safe submissions for a team. For simulation competitions this is every active (leaderboard-eligible) submission; for regular competitions it is the single submission currently on the public leaderboard.
• submit_to_competition — Submit a new competition submission.
• download_notebook_output — Download the output of a notebook.
• list_dataset_files — List the files in a dataset.
• download_hackathon_write_ups — Exports a CSV with the data from all submitted writeups after the hackathon has closed.
• get_competition — Get competition metadata.
• list_model_variations — List model variations.
• list_competition_data_tree_files — List the data files by directory for a competition.
• list_competition_topics — List discussion topics for a competition.
• get_writeup_by_slug — Get a writeup by its slug.
• get_resolved_writeup_links — Get resolved information for all links in a writeup, including download URLs, file summaries, and metadata for datasets, notebooks, YouTube videos, and external links.
• update_model_variation — Update a model variation.
• list_submission_episodes — List episodes for a submission in a simulation competition.
• download_competition_data_files — Download all data files for a competition.
• search_competitions — Search for competitions.
• save_notebook — Save a notebook and run it from top to bottom.
• get_writeup — Get a writeup by ID.
• get_forum — Get details about a discussion forum, including its name, description, and topic count.
• list_notebook_files — List the files in a notebook.
• search_notebooks — Search for notebooks.
• start_competition_submission_upload — Get a link to upload a competition submission file.
• create_notebook_session — Create a new notebook session.
• search_datasets — Search for datasets.
• list_notebook_session_output — List the output files of a notebook session.
• upload_dataset_file — Upload a file to a dataset.
• list_dataset_tree_files — List the files in a dataset by directory.
• get_hackathon_write_up — Get a single hackathon write-up submission by ID for a competition.
• list_topic_messages — List discussion messages within a competition topic.

About MCP

The Model Context Protocol (MCP) is an open protocol Anthropic introduced for connecting LLM-based agents to external tools and data sources. Providers publish MCP servers that expose their API surface as structured, discoverable tools — an MCP-compatible client (Claude Desktop, Cursor, Cline, Continue, etc.) can connect to the server and call its tools without any per-provider integration code.

Browse every MCP server on the APIs.io network or compare with the broader Agent Skill surfaces of the same providers.

Work with this as data

Every MCP server here is available over the APIs.io API and to AI agents over MCP.

curl "https://apis.io/api/v1/mcp/kaggle-mcp-server"

curl "https://apis.io/api/v1/mcp?limit=25"