ai-supply

The world's most popular computer vision library with Python bindings — image processing, video, and ML pipelines.

The largest collection of pretrained image models for PyTorch — ViT, ConvNeXt, EfficientNet, Swin, and 900+ more.

Apache-2.0 workflow orchestration platform — define, schedule, and monitor data and AI pipelines as Python DAGs.

Meta AI's promptable image segmentation model that can segment any object from a single click or bounding box.

Apache-licensed 7B language model from Mistral AI — beats Llama 2 13B on most benchmarks at half the size.

ProjectDiscovery's fast, template-driven vulnerability scanner with 9,000+ community templates for web apps, APIs, cloud, and AI service endpoints.

Hugging Face Transformers: state-of-the-art pre-trained models for NLP, vision, audio, and multimodal tasks.

The most popular open-source Yahoo Finance connector — historical prices, fundamentals, options chains, and real-time quotes as pandas DataFrames.

The leading framework for building LLM-powered applications and agents with chains, tools, memory, and retrieval.

Run large language models locally with a single command. OpenAI-compatible API for Llama, Mistral, Gemma, and more.

Aqua Security's comprehensive vulnerability scanner for containers, filesystems, Git repos, IaC, and SBOMs — now with AI/ML workload scanning.

AI pair programming tool in your terminal — chat with LLMs to edit code across your entire git repository.

The largest open prompt collection — thousands of curated system prompts for personas, coding, writing, education, and creative tasks. CC0 licensed.

Farama Foundation's MIT-licensed reinforcement learning toolkit — the standard interface for RL environments including Atari, MuJoCo, CartPole, and 100+ more.

MIT-licensed rule-based sentiment analysis tool tuned for social media, brand monitoring, and ad copy: compound scores with no model training required.

MIT-licensed SOTA English embedding model from BAAI — top MTEB leaderboard performer, commercial-friendly.

Meta AI's modular object detection platform supporting Mask R-CNN, Faster R-CNN, DETR, and panoptic segmentation.

Battle-tested deep learning TTS toolkit with XTTS, VITS, and 1000+ pretrained voices across 20+ languages.

Cross-browser automation library for agents: control Chromium, Firefox, and WebKit with a single reliable API.

384-dimensional sentence embeddings with tens of millions of downloads — fast, compact, and remarkably accurate for semantic search and clustering.

Microsoft's MIT-licensed 3.8B SLM — instruction-tuned, runs on CPU/mobile, punches far above its weight class.

Minimal keyword and keyphrase extraction using BERT embeddings — finds the most representative terms in any content for SEO and content strategy.

Microsoft's open-source PII detection and anonymization engine supporting 50+ entity types across text, images, and structured data.

Open-source AI code assistant VS Code and JetBrains extension — autocomplete, chat, and edit with any LLM.

DLR's MIT-licensed PyTorch implementations of PPO, SAC, TD3, DQN, A2C, and HER — battle-tested RL algorithms for game AI and simulation.

MIT-licensed comprehensive guide and prompt library — techniques, examples, and templates for every major LLM prompting method.

Official reference MCP servers: filesystem, git, fetch, memory, and more — the fastest way to give agents real-world tool access.

MIT-licensed Python NLP library with sentiment polarity/subjectivity scoring, spell correction, noun phrase extraction, and translation for copywriters.

Project MONAI's Apache-2.0 PyTorch framework for medical image segmentation, classification, and registration — the standard for AI radiology research.

Alibaba's Apache-2.0 7B instruction model — top multilingual performance, 128k context, strong at coding and math.

Roboflow's reusable computer vision utilities for annotation, tracking, and visualising detection model outputs.

OpenAI's open-source speech recognition model — accurate multilingual transcription and translation in one model.

OpenMMLab's comprehensive object detection toolbox with 40+ architectures and 300+ pretrained models.

Apache-2.0 toolkit for training intelligent game agents using deep RL and imitation learning directly inside Unity game environments.

Unified API gateway for 100+ LLM providers — call OpenAI, Anthropic, Gemini, Cohere, and local models with a single consistent interface.

Connect AI agents to the web: Browser Use lets LLMs browse, interact, and extract from any website autonomously.

Apache-licensed multi-cloud security assessment tool covering 500+ checks across AWS, Azure, GCP, and Kubernetes, including AI service misconfigurations.

The leading data framework for LLM apps — 150+ data loaders, RAG pipelines, and agent tools for connecting any data source to any model.

Self-hosted, open-source AI coding assistant — code completion, chat, and codebase indexing on your own infra.

Suno's transformer-based text-to-audio model that generates realistic speech, music, and sound effects from text.

Microsoft's universal document-to-Markdown converter: PDF, DOCX, PPTX, XLSX, HTML, images, audio, and ZIP — all to clean Markdown.

Apache-2.0 text embedding model with 8192-token context — the first open, auditable, long-context embedding.

Farama Foundation's MIT-licensed multi-agent RL environment library — 50+ cooperative and competitive games for training and evaluating MARL algorithms.

BERT pre-trained on 12GB of diverse English legal text for superior legal NLP: case law, contracts, legislation, and EU/US court documents.

MIT-licensed security toolkit by ProtectAI that sanitizes LLM prompts and responses — blocking prompt injection, toxic content, PII leakage, and secrets.

Build stateful, multi-actor agent workflows as directed graphs — cycles, branching, human-in-the-loop, and persistent state built in.

Apache-2.0 autonomous coding agent VS Code extension — plans, writes, runs, and debugs code end-to-end in your editor.

Facebook AI Similarity Search: blazing-fast library for efficient similarity search and clustering of dense vectors.

Role-based multi-agent orchestration framework — define a crew of AI agents with distinct roles, tools, and goals that collaborate autonomously.

Fast and accurate physics engine for robotics, biomechanics, and RL research, now open-sourced by DeepMind.

TypeScript toolkit for building AI-powered apps and agents: streaming UI, structured outputs, tool calls, multi-step agents.

The open-source AI-native vector database — store, query, and filter embeddings with a simple Python API. Perfect for RAG and semantic search.

Microsoft's open-source quant research and trading platform with ML/RL model support, factor mining, portfolio optimization, and backtesting.

Microsoft's multi-agent conversation framework — build teams of LLM agents that converse, code, debate, and solve complex tasks together.

MIT-licensed Python library computing 12 readability scores (Flesch, Gunning Fog, SMOG, Dale-Chall, etc.) for content optimization and SEO.

DeepMind's Apache-2.0 framework for game theory and RL research: 70+ games (Chess, Go, Poker, Hex), CFR, MCTS, deep RL, and Nash equilibrium solvers.

EleutherAI's MIT-licensed unified benchmark suite — the de-facto standard for evaluating language models across 200+ tasks.

High-performance vector database with filtering, payload storage, and a REST/gRPC API — built for production AI.

Open-set object detector that grabs any object by text description — no class list needed at inference time.

Hugging Face library for parameter-efficient fine-tuning — LoRA, QLoRA, IA3, and more. Fine-tune billion-parameter models on a single consumer GPU.

MIT-licensed 4.2M instruction dataset — GPT-4/3.5 augmented CoT traces that power top open-source fine-tunes.

AI4Finance Foundation's lightweight, open-source financial LLM fine-tuned on market news, earnings calls, and SEC filings for sentiment and forecasting.

Production-ready NLP pipeline framework for building search, RAG, and question-answering systems with any LLM.

Farama Foundation's Apache-2.0 fast gridworld RL environments for goal-conditioned, partially observable, and language-conditioned agent research.

Apache-2.0 fine-tuning framework — train any HuggingFace model with LoRA/QLoRA/full-fine-tune via a single YAML config.

Meta AI's music source separation model that splits any song into vocals, drums, bass, and other stems.

Microsoft's open-source SDK for building AI agents and skills in C#, Python, and Java — composable, enterprise-ready, and model-agnostic.

Allen AI's Apache-2.0 spaCy models for biomedical text with UMLS, MeSH, GO, HPO, and RxNorm entity linking and NER for genes, diseases, chemicals, and proteins.

MIT-licensed framework for evaluating LLMs and AI systems — build custom evals, run model comparisons, log results.

Expert-labeled dataset of 13,000+ annotations across 510 commercial contracts covering 41 legal clause types for contract review AI.

Real-time physics simulation for robotics and game AI using the Bullet physics engine with Python bindings.

Autonomous AI research agent that searches the web and produces detailed, cited research reports in minutes.

Stanford's framework for programming — not prompting — LLMs: compile, optimize, and auto-tune modular AI systems.

DKFZ's Apache-2.0 framework that automatically configures a U-Net for any medical image segmentation dataset — state-of-the-art results with zero manual tuning.

NVIDIA's open-source LLM vulnerability scanner that probes language models for prompt injection, jailbreaks, hallucinations, and more.

CC-BY-SA-3.0 instruction dataset of 15k human-written prompts — the first commercially licensed open instruction dataset.

MIT-licensed deep learning framework for drug discovery and computational biology — molecular property prediction, virtual screening, and ADMET modelling.

Microsoft's graph-based RAG: build knowledge graphs from documents for global, multi-hop reasoning beyond vector search.

Foundation model for monocular depth estimation with state-of-the-art relative and metric depth from a single image.

Official MIT-licensed Python SDK for the Model Context Protocol — build MCP servers and clients with full spec compliance.

Modern workflow orchestration — turn any Python function into a scheduled, observable, retriable data or AI pipeline with zero boilerplate.

Intelligent memory layer for AI agents — automatically extracts, stores, and retrieves relevant context across sessions.

OpenAI's MIT-licensed suite of 16 procedurally-generated 2D game environments for measuring generalization in reinforcement learning agents.

AI4Finance Foundation's framework for training RL agents to trade stocks, crypto, and forex with backtesting, paper trading, and live execution support.

MIT-licensed Python library for SEO, SEM, and digital marketing analysis: sitemap parsing, robots.txt analysis, keyword planning, and log file analysis.

Speaker diarisation and voice activity detection toolkit — who spoke when in any audio recording.

Microsoft's MIT-licensed GPT trained on 15M PubMed abstracts for biomedical relation extraction, question answering, and literature text generation.

Open-source document ingestion library — partition PDFs, HTML, DOCX, and 25+ formats into clean elements for RAG pipelines.

HuggingFace's few-shot text classification framework — match Legal-BERT accuracy with as few as 8 labeled contract examples, no prompt engineering.

Apache-2.0 LLM testing framework — pytest-style unit tests for RAG pipelines, chatbots, and LLM outputs.

All-in-one conversational AI toolkit for ASR, speaker recognition, speech enhancement, and language identification.

Git-like version control for ML datasets and pipelines — track experiments, reproduce results, and collaborate on data science projects.

MIT-licensed spaCy extension for clinical text processing — section detection, clinical concept extraction, negation, temporality, and UMLS entity linking.

Microsoft's open-source AI red-teaming toolkit for systematically finding risks in generative AI systems through automated adversarial probing.

Structured outputs for LLMs using Pydantic — patches any OpenAI-compatible client to return validated Python objects.

MIT-licensed BERT fine-tuned on MIMIC-III clinical notes for superior clinical embeddings — NER, relation extraction, and ICU outcome prediction.

Apache-2.0 RAG evaluation framework — faithfulness, answer relevancy, context recall, and more in one pip install.

End-to-end speech processing toolkit covering ASR, TTS, speech translation, enhancement, and speaker diarisation.

Pydantic's type-safe agent framework: build production agents with structured I/O, dependency injection, and full observability.

DeepMind's MuJoCo-based physics simulation library and continuous-control RL benchmark suite.

Apache-licensed Java/Python NLP toolkit with tokenization, sentence detection, NER, POS tagging, and chunking — production-ready for legal document pipelines.

Lightweight BSD-licensed Python library providing 80+ stock technical indicators (RSI, MACD, Bollinger Bands, ATR) as simple DataFrame column lookups.

LLM eval + red-teaming framework — test prompts and models against custom assertions, compare providers, and catch regressions in CI.

Block's Apache-2.0 on-machine AI developer agent — runs tasks end-to-end in your terminal, any LLM backend.

Guaranteed structured generation from LLMs: enforce JSON, regex, grammar, or Pydantic schemas at the token level.

Apache-2.0 unsupervised clinical NER and linking library that maps free-text clinical mentions to SNOMED CT, ICD-10, UMLS, and other biomedical ontologies.

Validate, fix, and filter LLM outputs — define structured schemas and safety rules, then automatically retry when outputs fail validation.

Google's Apache-2.0 TensorFlow library for quantitative finance — GPU-accelerated options pricing, rate models, Monte Carlo simulation, and autodiff Greeks.

State-of-the-art robot learning library from Hugging Face with pretrained policies for real and simulated manipulation tasks.

Apache-2.0 Python framework for building MCP servers fast — decorator-based, zero boilerplate, typed tool definitions.

ProtectAI's self-hardening prompt-injection detector using a multi-stage defence: heuristics, LLM analysis, and a vector canary database.

AI agent that retrieves, reads, and synthesises answers from scientific PDFs with citation-level accuracy.

Maintained fork of Quantopian's pyfolio providing comprehensive risk and return analytics, tear sheets, and performance attribution for quantitative strategies.

Python library for robot kinematics, dynamics, and trajectory planning with a familiar MATLAB-like API.

AI4Finance Foundation's multi-agent platform for financial tasks — market forecasting, report generation, portfolio management, and earnings analysis.

Multi-task benchmark for legal NLP with 7 datasets covering EURLEX classification, contract clause labeling, court judgement prediction, and more.

Open-source AI-powered app starter built with Next.js and E2B sandboxes — let users generate and run full-stack apps from natural language.

Modular robot simulation framework built on MuJoCo for benchmarking robot learning algorithms.

MIT-licensed Django email marketing platform: subscriber management, campaign builder, open/click tracking, and segmentation — fully self-hosted.

spaCy-based NLP pipeline for English legal text: named entity recognition for cases, legislation, and provisions, plus text categorization.

Interactive household simulation environment for embodied AI tasks including navigation and manipulation.

Self-hosted MIT-licensed SEO crawler and auditor: crawls websites, surfaces technical SEO issues, broken links, missing metadata, and duplicate content.

MIT-licensed NER models for German and multilingual legal documents, extracting courts, laws, citations, parties, and dates from case text.

MIT-licensed AI content platform for SEO blog posts, social copy, YouTube scripts, and podcast outlines grounded in real web research and competitor data.