Text-to-video web app from $29 a month (Starter, 2,000 credits) with an Oct 22, 2026 AI video agent; credits meter Veo 3.1 at 20 a second and avatars at 22.5 per 30 seconds (up 50% on Nov 5), fees are non-refundable and 17% of 685 Trustpilot reviews are one-star.
Strong, fairly priced retrieval models (Embed 5 from $0.08, Rerank 5 from $0.02 per 1M tokens) and private deployment; Command A+ is capped on Cohere's own API and $0.30/$1.50 via OpenRouter.
Python-first serverless GPUs and sandboxes billed per second (H100 $3.95/hr, $30/month free compute); watch idle-time billing, region and sandbox multipliers, and cold starts.
Capable autonomous agent for research, slides, sites and video from $20/month (4,000 credits); no per-task cost estimate, credits reset monthly and support complaints are common.
Accurate, low-priced speech-to-text ($0.15 to $0.21 per audio hour) with new $15-per-1M-character TTS, but products and models changed fast in 2026 and the Voice Agent API is being retired.
Polished docs-as-code hosting with a free 5-seat Starter plan, but every AI feature needs Pro at $450/mo annual ($540 monthly), and it has two disclosed security incidents.
Very fast GPT-OSS inference with a free plan and $0.15/$0.60 per 1M tokens; self-serve catalog is narrow, models retire fast and Llama is Enterprise-only.
Broadest self-serve menu of new open models and often a fast host; no free trial ($15 prepaid minimum), prompts stored by default and serverless models retire on 2-3 weeks' notice.
Bot-free AI meeting notes that users rate very highly (App Store US 4.96 from 14,634); free plan shows only 30 days, Business is $14 a user monthly, training is on until you opt out, and a CIPA consent class action was filed Jul 30, 2026.
Turns selected text into editable diagrams and decks; free with 500 credits a week and a watermark, Plus $12 a month ($9 yearly), Pro $30 ($22 yearly); refunds only within 24 or 48 hours and unused; little independent review evidence.
Quickest way to run thousands of AI models through one API, with fair per-output prices; owned by Cloudflare since Dec 1, 2025, prepaid credit expires after 12 months, dedicated GPUs cost more than rivals and support complaints are common.
One of the broadest AI creative apps, with 40+ models, its own Krea 3 editing model, an agent and an API; Pro is $35 a month ($21 yearly), but individual-plan units expire monthly, refunds are narrow and Krea's own pages disagree on prices.
The debugger skill is a productivity component designed to act as a debugging specialist for errors, test failures, and unexpected behavior. You can use it proactively when encountering any issues rel
This review page examines the official Stripe remote Model Context Protocol server for Claude Code, categorized under devtools in the public template library. The configuration targets developers work
This review page covers the debugging-strategies skill from the public template library, categorized under productivity. The component aims to transform debugging from frustrating guesswork into syste
This review page examines the Notion MCP server configuration component, which belongs to the productivity category within the public template library. The component is designated as the official Mode
The api-design-principles skill is a community component added on 2026-02-27 under the development category. It is categorized with a risk level of safe. The component focuses on mastering REST and Gr
This review page covers the Claude Code component named supabase from the category database. The component is an MCP server configuration designed to connect your Claude Code to Supabase using MCP. Th
This review covers the clean-code component from the public template library, categorized under development and marked with version 2.0 and critical priority. The component provides pragmatic coding s
This review page examines the playwright-mcp component found in a public template library. Classified under browser automation, it functions as a Model Context Protocol server that gives large languag