Best AI & LLM tools
4508 tools in the AI & LLM category across every type.
What these tools are for
- Hand your agent a structured way to call a model or an inference API.
- Compare outputs from two different providers on the same prompt.
- Wire retrieval, embeddings, or generation into an existing agent workflow.
How to compare them
- Provider coverage:
- Match the tool's supported providers and models to the ones your project uses.
- Cost exposure:
- Model-backed tools spend real API budget; check how usage and limits are configured.
- Maintenance signal:
- The provider landscape shifts weekly; prefer tools updated within the last few months.
- Config surface:
- Prefer tools whose environment-variable or config keys are documented plainly.
Each tool's page shows how to set it up: per-client config for MCP servers, a copyable install command for skills and plugins, and the source definition for agents and rules.
Skills
MCP Servers

AI orchestration platform with 100+ agents, swarm coordination, and self-learning memory for enterprise development.

PraisonAI
Build autonomous, self-reflecting AI agents and connect them to any MCP server in minutes.

Honcho
Memory infrastructure for stateful agents: store conversations, let Honcho reason in the background, query peer representations and context.

ai.klavis/strata
One MCP gateway that lets AI agents progressively use thousands of tools without blowing up the context window.

treg.to
OpenRouter for tools and data—compare catalog providers and call them from one hosted MCP endpoint.

Let AI agents build, script, and debug Unity games directly inside the Editor and in-game.
Let Claude and other LLMs watch videos locally with intelligent frame extraction and audio transcription.
Tamper-evident decision audit trail and outcome-weighted memory for AI trading agents.

Solana MCP by Vybe
Browse Solana wallets, trades, markets, and onchain data via live API calls through Claude and Cursor.

ai.smithery/ref-tools-ref-tools-mcp
Token-efficient AI documentation search and retrieval for coding agents and LLMs.

Autonomous spec-to-product coding agent with verified completion and 36 MCP tools.

Agent Swarm
Open-source operating system for AI work: orchestrate multi-agent teams with persistent memory, Docker workers, and integrations.

Vestige
Local-first memory for AI agents that traces failures back to their root cause, no cloud required.

Compartment
Durable, encrypted agentic memory that runs fully offline on your computer.

haiku.rag
Local-first agentic RAG with hybrid search, reranking, and multimodal document retrieval with citations.

Kronos
AI-powered stock research assistant for China A-shares, Hong Kong, and US markets with Kronos time-series forecasting
Adds research-backed Chain-Pattern Interrupts so AI agents pause, reflect, and avoid reasoning lock-in.

ai.smithery/docfork-mcp
Up-to-date library documentation for AI coding agents — now shut down (offline since June 14, 2026).
Cut AI context usage 60-90% with smart caching, compression, and a persistent knowledge graph—no account needed.

GrayMatter
Persistent agent memory as a single Go binary: knowledge graph, TUI observability, zero infra.

Long-term, multimodal memory for AI agents with character-aware profiles and mem0-compatible API

Entroly
Drop-in context compression and cost optimization for AI agents—reduce tokens without losing evidence.

AI-powered image generation with Google Gemini models—Flash speed meets 4K quality

Turn your team's Slack, Discord, Teams & Mattermost chats into a self-maintaining, queryable knowledge graph and wiki.

Ephemeral neural network swarm orchestration with WebAssembly acceleration and 84.8% SWE-Bench solve rate

Local RAG server for searching private documents without sending them to APIs

Context Engine
Compress logs, retrieval chunks, and code context into structured LLM-ready signal for agents and AI coding.

DeepSeek MCP Server
Official MCP server exposing DeepSeek V4 chat, FIM completion, model listing, and balance endpoints.

VisualTorch
Visualize PyTorch neural network architectures as static diagrams or animated GIF reveals in multiple styles.

Self-hosted, markdown-based knowledge-graph memory for AI agents — persistent, decaying, self-improving retrieval with zero cloud dependency.
Zero-trust, air-gapped Enterprise GraphRAG MCP server with offline, citation-grounded answers using tree-search and knowledge-graph reasoning.
Local-first persistent memory layer for AI agents with semantic search and multi-agent support.
Correction-first agent memory that learns how you think, with precision KPI tracking and 5 cognitive layers—local-only by default.

Tetrees MCP
Own, grow, and trade portable AI agent intelligence via TAIP/1 Packs and MCP.

Apollo MCP Server
Expose GraphQL operations as MCP tools for AI models to access and orchestrate your APIs.
Persistent cognitive memory for AI agents—semantic search, Hebbian learning, knowledge graphs, zero LLM calls.

io.github.ohad6k/emulo
Mine your AI coding session logs into a you.md profile so agents understand how you actually work.

DEPRECATED: Google AP2 agent-to-agent payment protocol server with no live endpoint.

Give AI agents real-time ML fraud risk scoring and decisioning via Sift.
Fiat-to-stablecoin onramp/offramp, wallets, and transfers for Latin American commerce.

DEPRECATED: HTTP micropayments protocol by Coinbase for agentic commerce (no live endpoint).
Ebbinghaus-based persistent memory for Claude—memories decay with time, strengthen on recall.

YourMemory
Persistent, self-improving memory for AI agents with biological decay and consolidation—your AI remembers across sessions.
Arbitrage detection and paper trading for prediction markets on Kalshi and Polymarket.

io.github.l33tdawg/sage
Persistent, consensus-validated institutional memory for AI agents with Byzantine-resilient infrastructure.

gflow-cli
CLI and MCP server to drive Google Veo video and Imagen image generation from your terminal or agent.

Payment layer for AI agents. Call thousands of paid APIs on Solana and Base with automatic x402 settlement.

io.github.plur-ai/plur
Local-first, open memory for AI agents—learn once, remember everywhere across Claude, Cursor, and other tools.
Persistent memory for AI agents using spreading activation — no vector DB, no API calls, fully offline.

Claude Code plugin + MCP server for ComfyUI: generate images, author workflows, manage models in natural language.

Access Alibaba Cloud DashScope Qwen models for chat, embeddings, and model discovery via MCP.

three.ws Audio
Text-to-speech, speech-to-text, audio-to-face lipsync, and motion-capture for 3D agents.

three.ws Autopilot
Set autopilot scopes and daily SOL spend caps, then propose, execute, and undo agent actions on-chain.

three.ws Brain
Route chat completions across multiple LLM providers via the three.ws multi-provider API.

IBM watsonx.ai Granite
IBM watsonx.ai Granite MCP server for chat, text generation, embeddings, and tokenization.

IBM Granite via x402
Pay-per-use IBM Granite AI models for chat, code, embeddings, and analysis via USDC blockchain payments.

IBM Granite x402 Remote
Pay-per-use IBM Granite AI models (chat, code, embeddings, forecasting) via x402 protocol on Base or Solana.

Metaplex Agent Registry
Deploy AI agents on-chain on Solana with 3D avatars and an EIP-8004 identity.

three.ws Naming
Resolve .sol names and reverse-lookup wallets for AI agents on Solana.

3D avatars, agent memory, and on-chain identity for AI agents—generate, customize, and embed interactive 3D models.
Plugins

agent-orchestration
Optimize multi-agent systems with workflow improvements and context management.

context-management
Persist and restore context for long-running conversations and multi-session workflows

dgx-spark-ops
NVIDIA DGX Spark environment setup and ML workload optimization for Grace Blackwell systems

llm-application-dev
Build LLM applications with LangGraph, RAG, vector search, and AI agent architectures

llm-finetuning
Eval-gated LLM fine-tuning with LoRA, preference optimization, and quantized export

machine-learning-ops
Automate ML model training, tuning, deployment, and experiment tracking workflows

math-olympiad
Solve competition math with adversarial verification to catch proof errors.

DeepEval
Add LLM evaluation, tracing, and dataset management to AI applications with DeepEval.

huggingface-skills
Access Hugging Face Hub operations and machine learning resources through Claude Code.

atomic-agents
Build, debug, and audit Atomic Agents Python applications with skills and diagnostic subagents.

NVIDIA Skills
Discover and integrate NVIDIA skills for GPU acceleration, CUDA, AI agents, and specialized workflows.

aws-agents
Build, deploy, and operate AI agents on AWS with Bedrock, MCP tools, and production hardening.

aws-startup-advisor
Personalized AWS guidance for startups: architecture, cost, security, migration, and AI stack optimization.

sagemaker-ai
Build, train, and deploy AI models with Amazon SageMaker directly in your coding assistant.

ai-engineer
Implement AI/ML features, language models, and intelligent automation for rapid deployment.
Implement AI ethics frameworks and governance policies for enterprise applications

angelos-symbo
Create and convert prompts using SYMBO symbolic notation system.

vision-specialist
Expert guidance on vision models, OCR, barcode detection, and visual AI tasks.

tavily
Real-time web search, extraction, and crawling for AI applications

langfuse
Tracing, prompt management, and evaluation tools for LLM applications via Langfuse.

pydantic-ai
Build production-grade AI agents with Pydantic AI tools, structured output, and streaming support

MLflow Skills
Instrument, trace, evaluate, and iterate on AI agents with MLflow.
Operate Oracle AI Data Platform Workbench in natural language—37-skill agent for Spark SQL, lakehouse ops, pipelines, and AI agents.

togetherai-skills
Agent skills for Together AI inference, training, embeddings, and multimodal processing.

north-star
System prompt that overrides RLHF defaults: removes confirmation-seeking, scarcity thinking, and best-practice ceilings.

datarobot-agent-skills
DataRobot skills for AI/ML workflows including model training, deployment, and monitoring.

AWS Startup Advisor
Personalized AWS guidance for startups: architecture, cost, security, migration, and AI stack rewrites.

anthropic-docs
Auto-updated reference skills for Claude, Anthropic API, MCP, and platform features.

xros
Executable research OS: compile investigations into verifiable methodology specs and run multi-agent verification workflows.

boltz
Predict protein structures, screen molecules, and design binders using Boltz computational biology models.
pixeltable
Build multimodal AI applications with tables, embeddings, and 25+ AI provider integrations.
Agents
Elite context engineering specialist for dynamic multi-agent AI orchestration, vector databases, and intelligent memory systems.

ai-engineer
Design, build, and optimize end-to-end AI systems from model selection through production deployment.

ml-data-expert
Expert ML & Data Science agent for building end-to-end machine learning pipelines with Python.

ai-engineer
Design, build, and optimize LLM-powered applications, RAG systems, and agentic workflows for production.

langchain-expert
Expert LangChain pipeline construction, document processing, and optimization

ai-engineer
Build production-ready LLM applications, RAG systems, and intelligent agents with enterprise-grade architecture.

data-scientist
Senior data scientist for exploratory analysis, predictive modeling, and statistical insights from datasets.

ml-engineer
End-to-end ML lifecycle management: from model development to production deployment, monitoring, and automated retraining.

openai-api-expert
Expert guidance on OpenAI API integration, authentication, optimization, and best practices.
Elite context engineering specialist for dynamic management, vector databases, and multi-agent AI orchestration.

llm-architect
Design and deploy production LLM systems with optimized inference, fine-tuning, RAG, and safety mechanisms.

prompt-engineer
Master prompt engineer architecting sophisticated LLM interactions and optimizing AI system performance.

pytorch-expert
Expert in building, training, and optimizing PyTorch deep learning models.

data-scientist
Expert data scientist for advanced analytics, machine learning, statistical modeling, and data-driven business insights.

Deploy, optimize, and serve ML models at scale with production-grade infrastructure and real-time performance.

scikit-learn-expert
Master scikit-learn for machine learning: model selection, feature engineering, and hyperparameter tuning.

eval-judge
LLM judge for plugin skill quality assessment using anchored rubrics across 4 dimensions.

mcp-developer
Build, debug, and optimize Model Context Protocol servers and clients connecting AI systems to external tools and data sources.

tensorflow-expert
Expert TensorFlow development: build, optimize, and deploy ML models at scale.

image-generator
Executes image generation requests with minimal overhead.

ml-engineer
Build production ML systems with automated pipelines, model serving, and continuous monitoring.

llm-finetuning-architect
Fine-tuning strategist who gates training on eval baselines and routes to the right method—often recommending against training entirely.

mlops-engineer
Design and implement production-grade ML infrastructure, CI/CD pipelines, and operational monitoring systems.

Independent eval gatekeeper: builds golden sets and graders before training, gates checkpoints after.

nlp-engineer
Build production NLP systems with transformer models, text pipelines, and multilingual support.

Executes fine-tuning lifecycle: dataset prep, training launch, monitoring, and model export per training brief.

prompt-engineer
Design, optimize, and test production LLM prompts for maximum effectiveness and cost efficiency.

ml-engineer
Build and deploy production ML systems with PyTorch, TensorFlow, and modern ML infrastructure.

quant-analyst
Develop quantitative trading strategies, financial models, and risk analytics with mathematical rigor and backtesting validation.

mlops-engineer
Build scalable ML infrastructure, pipelines, and production systems with MLflow, Kubeflow, and cloud MLOps tools.

Design, train, and deploy reinforcement learning agents for robotics, gaming, and autonomous decision-making systems.

prompt-engineer
Expert prompt engineer optimizing LLM performance through advanced techniques like chain-of-thought, constitutional AI, and production-ready systems.

quant-analyst
Build financial models, backtest trading strategies, and analyze market data with risk metrics and portfolio optimization.

vector-database-engineer
Expert vector database design, embedding optimization, and semantic search for RAG and similarity systems.
Rules
AutoML and hyperparameter optimization rules for Python ML projects using Ray Tune, Optuna, PyCaret, and time-series libraries
Design patterns and best practices for building agents, tools, and workflows with Google's Agent Development Kit.
Python LLM & ML development guide with modern tooling, type safety, and workflow best practices.
PyTorch and scikit-learn for chemistry ML with best practices for molecular data.
SQL-callable AI functions and hybrid vector+keyword search for RAG within Snowflake.
TensorFlow and deep learning best practices for building, training, and deploying neural networks
TypeScript LLM development with functional programming, multi-provider architecture, and composable APIs.
































