A headless API and hosted MCP Server that gives AI agents (Claude, ChatGPT) the ability to analyze what is working on TikTok, Instagram Reels, and YouTube Shorts. It goes beyond captions: it reads the video itself to extract hooks, CTAs, formats, on-screen text, and brand safety, then scores how far a video outperforms its creator’s baseline. It can run a Research on a niche (returns themes, winning hooks, sentiment, and recommended angles with lift computed in code), search and analyze individual videos, or set up Trackers on creators, hashtags, sounds, and Meta/TikTok ad libraries with Change Reports. Pricing is two-track: API/MCP is pure pay-as-you-go (prepaid credits, $1 = 1,000 credits**, minimum $10 top-up; new accounts get 1,000 free credits). A Research costs 1,036 credits (~$1.04) for up to 50 videos; video lookup is 3 credits, intelligence analysis is 23 credits. A separate **Web platform subscription** starts at **$99/mo (Starter, 10,000 credits) with a 7-day trial, though third-party sources list conflicting entry prices ($149/mo). Critical caveats: it is built for developers/agents (no rich visual dashboard), Capterra shows 0 user reviews (independent validation is thin), and credit consumption can scale quickly if Research is run frequently without estimate checks.
Pros & Cons:
Pros: The video-first analysis is a genuine differentiator—it reads hooks, CTAs, and on-screen text from the video itself, not just the caption, which means it can find untagged brand mentions and analyze visual trends that text-based social listening tools miss ; the MCP Server and pay-per-credit API make it accessible to AI agents without seat contracts or platform subscriptions, and estimate=true lets developers quote a cost before spending ; the Research output is structured for decision-making (themes, verbatim hooks with templates, winning formulas, audience gaps from comments) rather than raw data dumps ; Trackers on ad libraries (Meta/TikTok) and creators provide ongoing monitoring without manual sweeps.
Cons: Pricing is split across two models (API credits vs. Web subscription), and third-party records conflict ($99 vs. $149/mo), making it hard for non-developers to understand the actual cost without checking the official pricing page ; it is not a visual dashboard—the value is in the API/MCP output, so marketers expecting a Hootsuite-style interface will be disappointed; independent validation is essentially absent—Capterra shows 0 user reviews, and most evidence comes from ShortsIntel’s own docs and directory listings ; credit burn can be unpredictable—a single Research costs ~$1, and analyzing hundreds of videos without estimate checks could deplete a balance quickly.
Who is it for?:Developers and AI builders who want to give their agents (Claude, ChatGPT, Cursor) the ability to research short-form video trends or monitor brand mentions on TikTok/Reels/Shorts via MCP; B2C growth teams that want to find untagged creator mentions or analyze which hooks/formats are working, and are comfortable using an API or a lean web interface; agencies/analysts who need to track competitors, hashtags, sounds, or ad creatives programmatically. Not suited for marketers seeking a polished, all-in-one social media management dashboard (Hootsuite/Sprout are more appropriate), anyone needing extensive independent user reviews before adopting, or users who want unlimited usage without monitoring credit consumption.