Retrieve available chat sources, knowledge bases, and AI models in one call to configure chat and search.
Apache 2.0
606,731 tools. Updated 2026-09-24 10:29
"Reliable sources for use in training language models" matching MCP tools:
- Run data tests on models, sources, snapshots, and seeds, and execute unit tests on SQL models to ensure data quality and accuracy in your dbt projects.Apache 2.0
- List all available models in Open WebUI, including Ollama local, OpenAI, and custom pipe models, with IDs, names, and sources.MIT
- List trained models with type and metadata to review training scores and decide next steps.MIT
- Inspect a training CSV and generate a plain-language modeling plan before training. It identifies predictive columns, flags target leakage and ID-like fields, and suggests log-target usage, then asks for confirmation.-
- Create professional PowerPoint presentations from natural language descriptions. Ideal for pitch decks, training materials, or business reviews.MIT
Matching MCP Servers
- FlicenseNot gradedqualityCmaintenanceRead-only MCP server exposing homelab data sources (files, mail, calls) to Cloudflare OS, enabling read-only access to documents, emails, and call records.-
- FlicenseAqualityBmaintenanceEnables searching, comparing, and inspecting AI models by pricing, context window, and capabilities via the models.dev catalog.8-
Matching MCP Connectors
Browser support for web features, live from caniuse. From which version, and is it safe to ship?
61 verified financial models and a household planning engine: 20 MCP tools, 6 keyless, specs cited.
- Ask any DeFi question in natural language and receive answers synthesized from multiple data sources.MIT
- Download approved GGUF models into a local directory for offline transcription. Installs recommended English, Chinese, and language-detection models from the built-in allowlist.MIT
- Retrieve trending datasets from HuggingFace. Filter by task and sort by downloads, likes, or modified to find datasets for training and fine-tuning models.MIT
- List SageMaker endpoints, training jobs, notebook instances, and models in a specified AWS region to enable Well-Architected validation.Apache 2.0
- Import pre-loaded training data from Intigriti, PortSwigger, and other sources to enhance security testing skills for bug bounty hunting.MIT
- Get AI-recommended mental models for problem-solving by describing your challenge in natural language.Apache 2.0
- Retrieve endpoint, payload, async settings, and code examples for APIAny models by model ID, type, or language. Filter by model, type, language, or limit results.MIT
- Start PyLaia HTR model training for a collection, applying Transkribus UI-default preprocessing and parameters to achieve accurate models. Override settings as needed.Functional Source , Version 1.1, MIT Future
- Create a computer vision model by describing what to detect in natural language. No training data required—start building with Roboflow Rapid.-
- Plan model experiments by selecting live models, Cookbook recipes, and pinned Hugging Face candidates without running training, so you can compare viable options before committing resources.Apache 2.0
- List vision-language models with their prompts and capabilities. Discover available models to select the right one for your image tasks.-
- Submit a training job for a model version and monitor its status until TRAINING_COMPLETED or TRAINING_FAILED.MIT
- Retrieve training datasets in your account to locate dataset IDs and names for use in AI Toolkit configurations. Optionally returns full dataset payloads.MIT
- Retrieve all valid news language codes to filter news searches by language. Use these codes as input values for news queries.MIT
- Create an empty training dataset in draft status, ready for file uploads and status polling before submitting a training job.MIT