Explore AI Tools

An AI mentor platform that builds a searchable knowledge base from one expert’s public videos, podcasts, articles, and posts, then answers your questions using only that person’s published work. Every point links back to the source—the exact moment in a video, the page for an article or episode—and if the person never covered your question, the mentor says so instead of inventing an answer. It remembers your context (stage, goals, numbers) across sessions. A remote MCP server lets Claude Code, Cursor, and other agents ask the same mentor before acting (e.g., checking pricing advice before building a pricing page); tools include list_personas, search, ask, get_topic, and timeline. Operated by a Berlin-based company; payments via Polar; data hosted in EU/US; conversations are not used to train models. Pricing: Starter $19/mo (400 questions), Pro $29/mo (1,200), Max $59/mo (4,000); annual billing costs 10 months for 12. A third-party directory listed $12.99/$24.99 in June 2026, suggesting a price change—confirm at signup. Note: English-only; public mentors currently limited to business/startup topics (Arvid Kahl, Rob Walling, Marc Lou); mentors are not endorsed by the people they are built from; free access exists but its form varies across sources (3 questions per mentor page, or a lifetime 10-message free tier)—verify on the pricing page; no verified third-party user reviews on Trustpilot/G2/Capterra as of listing.

A short-form video intelligence platform with two distinct sides. The developer-facing side (dev.virlo.ai) is a pay-per-request data API and MCP server covering TikTok, YouTube Shorts, and Instagram Reels: Content Research Agents that collect videos by niche with 79 AI labels (topic, hook, format, tone, brand safety, AI use), creator/sound/hashtag lookups, tracking with daily AI reports, trends/digests, and an MCP server (https://dev.virlo.ai/api/mcp/mcp) for Claude Code and Cursor. Billing is prepaid (1 credit = $0.01, errors free); Trends Digest costs $0.25, creator lookup $0.50, agent run $0.50, raw video data $0.01. The consumer-facing side is an AI content creation tool (script/video/audio/image generation via Veo 3, Runway, Sora 2, ElevenLabs), sold via subscription tiers—but the intro subscription is data-only, and content-creation credits require an additional purchase or a higher tier (a point of complaint on Trustpilot). G2: 4.9/5 (6 reviews, all small businesses, Apr–Jun 2026); Trustpilot: mixed (two 1-star reviews about subscription not including creation credits, one 5-star review). Note: the two product lines have different pricing and access models—verify which one you\'re evaluating.

A read-only MCP server and web dashboard for demand research across Reddit, X, Hacker News, and Product Hunt. Connect it to Claude Code, Cursor, or Claude Desktop, and your agent can search discussions, find threads where people ask for a tool or complain about a product, read full comment trees, and discover subreddits with fit scores and self-promotion rules—every result carries a permalink. Agents can save post/reply drafts in your voice, but nothing posts from the terminal; every draft waits for your approval in the web app, with the same cooldowns and checks before publishing. Pricing: Free (10 requests/day, no card), Starter $9/mo (100 requests/day, 3,000/month), Pro $19/mo (unlimited requests, unlimited watches). One tool call = one request regardless of how many results it returns. Note: read-only by design—no posting, voting, replying, or DMs; dependent on Reddit/X/HN/Product Hunt APIs (availability may change); limited independent user reviews (one positive comment on StartupBase, no Trustpilot/G2 ratings as of listing).

An AI tool that turns PDFs, notes, and study material into short, viral-style “brainrot” videos for TikTok, Reels, and Shorts. Upload a document, pick a character and video style (Minecraft gameplay or meme images), and it generates a narrated video that summarizes the content using internet slang, meme references, and optional AI voices (including a LeBron James-style voice). Built by two students (Columbia and Caltech) who previously ran a TikTok account with 15M+ views. Their 23 accounts have collectively pulled 100M+ views, 342K shares, and 220K bookmarks. Pricing: Free ($0, limited generations), Chat ($5/mo, removes watermark), Rizzler ($20/mo, dedicated servers, faster processing, better quality, large documents). Note: the product was originally a TikTok trend-analysis tool and pivoted after its “brainrot translator” feature went viral—early directory listings may be outdated. No verified third-party user reviews on Trustpilot/G2/Capterra as of listing. Not to be confused with the MemeNow meme generator app on Android.

A unified social media data and video analysis API for developers and AI agents. Extracts transcripts, AI summaries, comments, stats, channel data, and downloadable files from YouTube, TikTok, Instagram, Facebook, Twitter/X, and LinkedIn. Coverage includes video transcripts (2 credits/started minute), AI summaries (2 credits/request), comments/search/listings (1 credit per 50 results for most platforms, 1 per 20 for Instagram), and downloads (1 credit/started minute up to 480p, 4 credits for 720p/1080p). Failed requests cost 0 credits. Offers Agent Skills (npx skills add SocialKit-API/skills --all) for Claude Code, Cursor, and Codex, with MCP optional. Note: developer/agent tool, not an end-user app; platform coverage varies—Instagram search returns only one batch of up to 12 reels with no pagination, LinkedIn only supports publicly accessible profiles, and Facebook lacks bulk/download endpoints. Rate limits scale with effective credit allowance (40 req/min below 2,000 credits up to 600 req/min above 50,000). Subscription plan pricing is not fully documented in the API docs; verify on the official site. Not to be confused with socialk.it, a separate social media scheduling product.

A unified social media publishing API for developers and AI agents. Integrate once to publish, schedule, and manage comments/DMs across LinkedIn, X/Twitter, Instagram, Facebook, TikTok, YouTube, Threads, Pinterest, Reddit, Bluesky, Google Business Profile, Telegram, and WhatsApp. Handles OAuth tokens, Meta App Review, webhook subscriptions, and platform-specific quirks (e.g., Instagram\'s 24-hour DM window, TikTok\'s mandatory creator-info query, YouTube\'s quota system) on its side. Ships an official MCP server (postproxy-mcp) for Claude/Cursor integration. Pricing: Free (2 profile groups, 10 posts/month), Build $17/mo (10 groups, 120 posts), Grow $49/mo (20 groups, unlimited posts), Scale $99/mo (50 groups, SSO), Enterprise $699/mo (500 groups). Cross-posting the same content to multiple platforms counts as 1 post. Note: developer/agent tool, not an end-user app; X has a default hard cap of 20 posts/day (can be raised on request; BYO X credentials bypass this shared quota); capabilities vary by platform; ScamAdviser trust score 47/100 (low traffic, but flagged safe by DNSFilter).

An open-source (Apache 2.0) HTML-to-video rendering framework by HeyGen, designed for AI coding agents. You write video compositions as HTML/CSS with data attributes and GSAP animations, preview in the browser, and render deterministic MP4 via headless Chrome + FFmpeg. Installable as agent skills for Claude Code, Codex, Cursor, etc.; a Studio app and hosted connector let agents build and render videos in a chat interface. Best suited for information-dense content: data visualizations, UI demos, product explainers, tutorial overlays, motion graphics. It does NOT generate cinematic footage or realistic people—for that, use generative video models (Runway, Pika, Veo). Note: there are multiple unrelated products named \"HyperFrames\" (hyper-frames.com, hyperframe.ai); this entry refers specifically to HeyGen\'s framework at hyperframes.studio / hyperframes.heygen.com. Free for local rendering; cloud rendering and HeyGen voices/avatars bill through HeyGen credits. Chinese text layout may require pre-rendering titles in design tools.

A free macOS menubar app that captures a screenshot and asks GPT-4o to analyze or explain it. Useful for debugging code errors, generating code from Figma designs, getting UI/UX feedback, analyzing charts, or solving math equations (LaTeX rendering supported). Built by Daniel Nguyen (also of BoltAI) as a native macOS app for Intel and Apple Silicon. Free to download; requires your own OpenAI API key (BYOK), with costs based on OpenAI usage. API key stored in Apple Keychain; developer does not store or analyze your content. macOS 12+ only.

Agentic Document Extraction (ADE) by LandingAI, Andrew Ng\'s enterprise AI company. A vision-first API that parses, splits, and extracts structured data from complex documents—PDFs, scans, images, dense tables, multi-page forms—with layout-aware Markdown output, precise page/coordinate citations for every block, and confidence scores for human review. Supports schema-first field extraction (flat, nested, arrays, multi-table), multi-document splitting and classification, and audit-ready bounding-box citations per value. Built for RAG pipelines, reconciliation, compliance checks, and document archives. SOC 2 Type II, GDPR, HIPAA; cloud, on-premises, or virtual private deployment; zero data retention option. Claims 1B+ documents processed and <2s processing time. 50+ enterprise customers including Fortune 100 financial services, Eolas Medical, GreenLite, and Serimag. Pricing based on document volume; contact sales for a quote.

Open-source local AI assistant

High-throughput LLM serving

LLM inference in C/C++

Local LLM desktop app

Run LLMs locally

Serverless AI compute

Serverless AI compute

Vector database

Run ML models via API

AI model hub and inference

AI workflow automation

Workflow automation with AI

Multi-agent conversation framework

ulti-agent orchestration