AI Apps Directory

Open compute marketplace and RL training stack for agentic models.

Prime Intellect aggregates GPUs from 50+ providers into an on-demand marketplace (1–256 GPUs with SLURM or Kubernetes orchestration) and pairs it with an open reinforcement-learning stack: the prime-rl async training framework, the Verifiers environment library, and a hub of 2,500+ community RL environments. Its INTELLECT model series demonstrates the stack at frontier scale, with weights and training recipes released openly.

AI code review with full-codebase context on every pull request.

Greptile builds a graph index of your repository and runs a swarm of agents over each pull request, catching cross-file logic bugs, security risks, and style violations that diff-only reviewers miss. It learns team standards from past review comments, enforces custom rules written in plain English, and reviews PRs on GitHub and GitLab, with TREX generating and running tests per PR.

Training data, evaluations, and enterprise GenAI from the data-labeling giant.

Scale supplies the human-annotated training data behind most frontier AI labs through its Data Engine, spanning labeling, RLHF, and expert red-teaming. On top of the data business it runs evaluation leaderboards, an enterprise GenAI platform, and Donovan, its platform for the US public sector.

Lifelike enterprise voice AI agents for customer service calls.

PolyAI's Agentic Dialog Platform builds voice agents that resolve customer service calls for large enterprises in banking, healthcare, hospitality, telecom, and retail. Agent Studio gives non-technical teams a no-code builder while the ADK exposes developer APIs, with SOC 2, HIPAA, GDPR, and PCI DSS compliance built in. Its proprietary Raven model is trained on over a billion enterprise conversations.

AI autocomplete and writing assistance inline across Gmail, Docs, and the web.

Compose AI is a browser extension that adds AI-powered autocomplete and text generation directly into the fields you already type in — Gmail, Google Docs, and most web apps. It finishes sentences as you write, drafts full emails from a one-line instruction, rephrases text, and supports custom shortcuts to expand snippets. The core extension is free, with a premium tier for personalization features.

Turn your text into editable diagrams, charts, and infographics — no prompting.

Napkin AI generates business visuals directly from text you already wrote: paste a paragraph, click, and it produces diagrams, flowcharts, scenes, and infographics with no prompt required. The AI reads context and relationships and offers multiple visual options you can recolor, restyle, and edit, then export as PPT, PNG, PDF, or SVG. It supports 60+ languages and real-time team collaboration.

Generate royalty-free music from text, built on a fully licensed catalog.

Loudly is an AI music platform that generates studio-quality, royalty-free tracks from a text prompt or a chosen genre and mood in seconds. It adds a remix studio, stem splitter, and sample-pack generator, plus a distribution add-on that ships finished tracks to Spotify, Apple Music, YouTube, and 40+ platforms. The Berlin company stresses that its model is trained only on its own consent-licensed music dataset, not scraped audio.

Record studio-quality podcasts and video remotely, then edit with AI.

Riverside is a browser-based studio that records each participant locally in separate tracks — up to 4K video and uncompressed audio — so quality isn't tied to anyone's internet connection. Its AI layer handles text-based editing, filler-word and background-noise removal, automatic clip generation, show notes, captions, and translation/dubbing into 30+ languages. A 2025 chat-based editor lets creators cut and repurpose footage by talking to an agent.

AI meeting notes that turn calls into finished decks, docs, and action items.

Supernormal joins your Google Meet, Zoom, and Microsoft Teams calls to record, transcribe, and write structured notes with summaries and assigned action items the moment a meeting ends. Beyond notes, it now drafts post-meeting deliverables — slides, spreadsheets, and documents — from the meeting context plus connected files. Notes push automatically to Notion, Slack, Salesforce, HubSpot, and Google Drive.

Community directory for discovering Model Context Protocol servers.

mcp.so is a community-driven directory that collects and organizes third-party MCP (Model Context Protocol) servers so developers can browse, share, and integrate external tools for AI apps. It ranks servers by real call volume with a "Call Ranking" leaderboard and offers Featured, Latest, Official, and Hosted filters. The project is open-source and still in preview.

GPT-powered AI tutor for conversational language practice.

Talkpal is an AI language-learning app that lets you hold open-ended voice and text conversations with a GPT-powered tutor across 80+ languages. It adapts to your level, corrects grammar and pronunciation in real time, and offers roleplays, debates, and topic-based practice. A free tier allows limited daily practice, with Premium removing limits.

An AI assistant that lives inside your messaging apps.

Poke is a personal AI assistant from Interaction that works through the chat apps you already use — Apple Messages, WhatsApp, Telegram, and RCS — rather than its own app. It connects to email, calendar, and 30+ services to draft replies, manage your inbox, schedule, track health, and run proactive automations called "recipes." Most non-real-time use is free; for heavier use you negotiate a monthly price with its AI before subscribing.

ByteDance's free AI assistant for chat, writing, and documents.

Doubao (豆包) is ByteDance's consumer AI assistant for everyday Q&A, content creation, document processing, translation, coding help, and multimodal interaction. Core features — dialogue, basic creation, document summarization, and translation — are free with no ads or paywalls. It runs on ByteDance's own Doubao models and is the most-used AI chatbot in China.

OpenAI's coding agent across terminal, IDE, web, and ChatGPT.

Codex is OpenAI's agentic coding tool that reads, edits, and runs code on your behalf. It spans a local terminal client (the open-source Codex CLI), a VS Code extension, a cloud agent that executes tasks in isolated containers with full repo access, and a surface inside ChatGPT. It runs on OpenAI's Codex-tuned models and can parallelize multiple agent tasks at once.

Open-source annotation platform for vision AI datasets.

Data-labeling suite for images, video, and 3D: bounding boxes, polygons, segmentation, keypoints, and object tracking, with AI-assisted labeling via SAM and custom models through its API and SDK. Ships as the MIT-licensed Community edition to self-host, the hosted CVAT Online with free and paid plans, or a self-hosted Enterprise tier.

AI-native collaborative notebook for data teams.

Collaborative data notebook positioned as a drop-in Jupyter replacement: real-time multiplayer editing, SQL + Python + R blocks, warehouse integrations (Snowflake, BigQuery, Redshift), and notebooks that deploy as data apps and dashboards. Deepnote AI generates, fixes, and explains code, and an agent can run analyses autonomously.

Energy-first AI cloud and gigawatt-scale AI data centers.

Vertically integrated AI infrastructure company: Crusoe Cloud offers NVIDIA and AMD GPU clusters with managed Kubernetes and managed inference, while its data-center arm develops and powers gigawatt-scale 'AI factories' — including OpenAI's 1.2 GW Stargate campus in Abilene, Texas, which Crusoe built and co-owns.

Full-stack AI cloud for training and inference at scale.

AI-native cloud built around large NVIDIA GPU clusters — bare-metal and virtualized H100 through Blackwell hardware with InfiniBand networking, managed Slurm and Kubernetes, high-speed storage, and MLOps tooling. Its Token Factory layer adds managed per-token inference for open-weight models. Microsoft signed a five-year capacity deal with Nebius worth up to $19.4B.

Always-listening AI companion pendant that texts you about your day.

Friend is a wearable AI companion: a microphone pendant that listens to your day over Bluetooth and sends proactive, personality-driven comments and conversation through its iOS app. The $129 device is a one-time purchase with no subscription, and a free web chatbot lives at friend.com. It does not speak aloud — replies arrive as text messages.

Voice cloning, audio watermarking, and deepfake detection in one platform.

Resemble AI spans both sides of synthetic voice: generating it and policing it. The platform offers voice cloning and text-to-speech built on its Chatterbox models, real-time audio watermarking, and Detect, a multimodal deepfake detector covering audio, image, and video. It deploys in the cloud or fully on-premises for regulated environments.

AI voice generator studio with 200+ voices, dubbing, and a low-latency TTS API.

Murf AI is a text-to-speech platform pairing a studio editor — 200+ voices across 35+ languages, voice cloning, dubbing, and a voice changer — with developer APIs. Its Gen 2 speech model focuses on pronunciation accuracy and granular voice controls, while the Falcon API targets sub-130ms latency for real-time voice agents. Integrations include Canva, PowerPoint, and Google Slides.

The Superintelligence Cloud — on-demand GPUs, 1-Click Clusters, and superclusters.

Lambda is a GPU cloud for AI training and inference, spanning on-demand HGX B200 and H100 instances, self-serve 1-Click Clusters, and single-tenant superclusters built on NVIDIA's latest generations. A GPU specialist since 2012, it sells compute by the hour without long-term hyperscaler contracts and co-engineers large deployments with NVIDIA.

The AI hyperscaler — GPU cloud built for large-scale training and inference.

CoreWeave is a purpose-built AI cloud renting large-scale NVIDIA GPU capacity for training and inference, layered with managed Kubernetes, AI object storage, and Mission Control observability. Public on Nasdaq since March 2025, it counts most leading AI labs — including OpenAI, Meta, and Anthropic — among its customers, with a contracted revenue backlog reported near $100B in 2026.

Native vector storage and querying in S3 — serverless, billion-vector scale.

Purpose-built vector storage inside Amazon S3: store and query up to two billion vectors per index across thousands of indexes per vector bucket, with S3's durability and elasticity and no clusters to manage. Frequent queries return in around 100ms and infrequent ones in under a second, and it plugs directly into Amazon Bedrock Knowledge Bases for RAG. You pay only for storage and queries.