Browse & Compare AI Tools

ChatGPT (OpenAI, San Francisco; launched Nov 2022, 800M+ WAU / ~1.2B MAU est. 2026) = the generalist AI surface​ — chat + GPT-5.6 Sol/Terra/Luna routing + Codex agent + ChatGPT Work + Deep Research + Images 2.0 + Voice (GPT-Live-1) + Memory/Projects/Custom GPTs + 60+ connectors. Not Claude (writing/coding-first), not Gemini (Workspace-native), not Perplexity (cite-first). 

Claude (Anthropic, San Francisco; public Mar 2023, Fable 5 / Opus 5 / Sonnet 5 / Haiku 4.5 lineup 2026) = the writing-/coding-/long-doc-safe generalist​ built on Constitutional AI — chat + Claude Code + Claude Cowork + Artifacts + Projects + Research + 60+ connectors (Gmail/Drive/Slack/M365/GB). Not ChatGPT (broader multimodal/ecosystem), not Gemini (Workspace-native), not Character.AI (RP).

Gemini = the Google-shaped AI surface: open gemini.google.com / iOS / Android / Chrome side-panel → chat in 40+ langs with Gemini 3.6 Flash​ (free default) or bump to 3.1 Pro / 3.7 Flash (via Spark)​ → paste 1M-token PDF/codebase → Deep Research writes cited report → NotebookLM (500 notebooks, 300 sources each, AI Pro+) for source-bound Q&A + Audio Overview → Gemini Live voice interrupt → Gems (custom mini-assistants) + Canvas (prompt→app/quiz/infographic) → Nano Banana 2 image gen/edit in-thread → Veo 3.1 text/image→8s clip with synced dialogue/audio (Lite on Pro, 1080p on Ultra) → Flow filmmaking suite → Workspace: draft in Gmail, summarize Drive, build Slides, analyze Sheets, Meet recaps → Android default assistant (replaces Assistant, reads screen/SMS w/ consent) → Chrome Auto Browse (AI Pro+ multi-step form-fill, US-only Aug 2026) → Spark agent (AI Pro+ 160+ countries, Mariner deeper on $199.99 Ultra) → API via ai.google.dev (3.7 Flash $0.75/$3.75, 3.1 Pro $2/$12 per 1M, >200K doubles). Fits ai-productivity-data; pipeline_stage none (chat→deep research→notebooklm→workspace→spark→api). For Google-stack workers, long-doc/codebase ingestion, video-with-audio gen, Android users — default; for privacy-strict client work, Claude; for agentic coding, Claude/ChatGPT.

Grok = the real-time-social-native consumer assistant: grok.com + iOS/Android apps + X-integrated pane → ask anything with live web + X firehose (DeepSearch, 40+ sources, <1-min cited reports)​ → Big Brain extended reasoning → Grok Imagine​ (Aurora images + Imagine video 480p/720p w/ audio) → Voice mode → Grok Bot​ AI teammates (Heavy-only beta, each gets a cloud machine, bot-manages-bot, signs into inboxes/CRMs) → Grok Build​ coding CLI/TUI (Cursor-trained; GitHub Copilot dropped Grok Code Fast 1 on May 15 2026). Fits ai-productivity-data; pipeline_stage none (prompt→search→reason→generate→tool-call→act). For X-native marketers, trend/mood trackers, cost-sensitive API builders, and heavy agentic users, it's the only seat with native firehose; for deepest prose, safest guardrails, or regulated work, Claude/ChatGPT win.

Hugging Face = open-model infrastructure + community: discover Llama 4 / DeepSeek-R1 / Qwen3.5 / Flux / Whisper on Hub → transformers load in 3 lines → fine-tune via AutoTrain or peft → demo on Gradio Space (ZeroGPU free tier or paid GPU) → serve via Inference Providers (routed, per-token, no markup) or dedicated Endpoints (per-hour GPU) → govern in Enterprise Hub (SSO/SCIM/audit/resource groups/IP ranges) → script it all via hf_mcp_server (Jul 2026). Fits ai-productivity-data; pipeline_stage none (discover→load→fine-tune→demo→serve→govern). For ML engineers, researchers, indie devs shipping open-weight demos, and enterprises needing SSO over a model catalog, it's the default; for no-ops product teams wanting one API call and zero Hub literacy, Replicate/Together win.

HuggingChat = the open-weight chat playground: open chat.huggingface.co → guest or free HF login → pick Llama 4 70B / Qwen3 / Kimi K3 / DeepSeek-R1 / GLM-4.6 / Command A, or let Omni​ route per turn → web search via Exa MCP, file/PDF/image upload, image gen delegated to Z-Image-Turbo on ZeroGPU, CSV batch mode, custom system prompt, share link → same models callable via Inference API if you outgrow the UI. Fits ai-productivity-data; pipeline_stage none (open→pick/route→search/upload→answer→export). For devs/researchers A/B-ing open models before self-hosting, students wanting a $0 ChatGPT-class assistant, and privacy-minded users who want weights-inspectable, it's the default free seat; for polished voice/canvas/GPTs/agent store, ChatGPT wins.

Mistral Vibe (ex-Le Chat) = the EU-boundary assistant: chat.mistral.ai → pick Small 4 (free) / Large 3 / Medium 3.5 → Flash Answers (~1,000 tok/s), Think mode (Magistral CoT), web search with citations, document upload (PDF/CSV/Office), Canvas (doc/table/slide/code), Flux image gen, 100+ connectors (Notion/Slack/Linear/Drive/GitHub/Outlook) → Work Mode runs multi-step mail/calendar/inbox triage with explicit approval gates → Vibe for Code (Devstral-powered CLI + VS Code/JetBrains + remote sandbox agents, teleport session local↔remote) → same weights callable via La Plateforme API (Small 4 $0.15/$0.60, Large 3 $0.50/$1.50, Medium 3.5 $1.50/$7.50 per M). Fits ai-productivity-data; pipeline_stage none (chat→search→reason→act→code→deploy). For EU/GDPR-bound teams, multilingual (FR/DE/ES/IT) work, cost-sensitive Pro users, and devs wanting Devstral coding + self-host exit ramp, it's the jurisdiction play; for max frontier reasoning/voice/video/GPTs ecosystem, ChatGPT/Claude win.

Lumio AI (Multi‑Model Unified Workspace) 1. One‑line positioning Field Content Title Lumio AI (Multi‑Model Unified Workspace) Official URL https://lumioai.app One‑line positioning Lumio AI is a cost‑effective unified multi‑model AI workspace that consolidates leading LLMs including GPT‑5.5, Claude Opus, Gemini 3.1, Grok and DeepSeek within one web interface; delivers side‑by‑side model comparison, smart automatic model routing,…

Meta AI is the zero-install assistant already in your WhatsApp thread, IG search bar, FB feed, and Messenger — plus meta.ai web and a standalone iOS/Android app. It answers, drafts, summarizes, translates, does real-time web lookup, generates/edits images inline ("Imagine"), runs full-duplex voice on mobile and WhatsApp, answers camera/visual questions, and on Ray-Ban Meta glasses becomes a heads-up assistant. Engine is Llama 4 Scout​ (fast) / Maverick​ (default chat) with optional session memory; Behemoth​ is teacher-only, not shipped. Strategy is reach-not-revenue: free forever for casual users, monetization via ads + Meta One pilot + Llama API/enterprise, not seat licenses.

Microsoft Copilot = the AI that already lives inside the apps your company opens at 9am, not a chatbot you choose to open: free copilot.microsoft.com / Edge sidebar / Windows Win+C for web chat, voice, Think Deeper, Imagine image gen, shopping, vision-on-screen → sign into Microsoft 365 Premium ($19.99) for Copilot inside Word/Excel/PPT/Outlook/OneNote with higher limits + Researcher/Analyst agents → buy Microsoft 365 Copilot Business ($18–21/seat/yr-commit) or Enterprise ($30/seat/yr) on top of a qualifying M365 base license to ground in Graph, summarize Teams meetings, draft from SharePoint, triage inbox across calendar+meetings, build agents in Copilot Studio → Agent 365 / E7 Frontier Suite for governed multi-agent fleets. Fits ai-productivity-data; pipeline_stage none (chat→ground→draft→agent→govern). For M365-heavy orgs that want AI without new logins or connector plumbing, it's the path of least resistance; for best standalone reasoning/model quality, ChatGPT/Claude still win.

Perplexity = the research stack that starts where search ends: type a question → Sonar model fires dozens of real-time searches → reads full source docs → synthesizes with numbered inline citations you click-verify​ (reasoning chain visible) → follow-up threads keep context → Free 5 Pro/day, Pro $20 unlocks Deep Research (40-source reports ~4 min, 90%+ accuracy), model picker (GPT-5.6 / Claude Sonnet 5 / Sonar 2), file/image upload, image gen, Labs (code/data/asset generation), Pages (shareable formatted reports), Spaces (team curated research desks with custom instructions) → Brain (Max, 15 langs, self-improving wiki memory, +10pp memory benchmark, -6% dissatisfaction) → Perplexity Computer​ (Feb 25 2026, Firecracker microVM, orchestrates 20 models, Slack @computer + web/mobile, enterprise Snowflake/Salesforce/SharePoint/HubSpot/Datadog + SSO/SCIM/audit/SOC2/zero-retention) → Comet browser​ (Chromium, address bar = answer engine, sidecar reads current tab, agent fills forms/compares pricing/cross-tab summary, voice, ad/tracker blocker, Finance Hub, Model Council on Max; free Win/Mac/iOS/Android since Mar 18 2026, Comet Plus $5 publisher content). Fits ai-productivity-data; pipeline_stage none (query→search→synthesize→verify→report→agent-act). For journalists/investors/analysts/policy researchers needing auditable real-time intelligence, it's unmatched; for long-horizon academic PDF synthesis, Gemini 2.5 Pro deeper; for creative/open chat, ChatGPT/Claude.

Poe = the model-comparison cockpit, not a chatbot: sign in at poe.com / iOS / Android / Win / mac → pick Assistant (default router) or any bot → ask GPT-5.6-Sol, flip to Claude Opus 5, run Gemini 3 Pro on the same thread ("compare") → spawn a multi-bot chat (side-by-side grids) → build a custom bot (system prompt + model pin + optional Knowledge Base pre-May 2026 + monetization price) → publish to marketplace, earn per-message USD from others' point spend → Group Chat (Nov 2025, 200 participants, each spends own points, mention @GPT5 @Claude in-thread) → Poe Apps (Feb 2025, no-code canvas/app builder) → Script Bots (late 2025, Python workflows chaining models) → Poe API (Jul 2025, OpenAI-compatible, tool-calling for OpenAI/Anthropic/Google, 500 req/min, plugs into Cursor/Cline/n8n). Fits ai-productivity-data; pipeline_stage none (pick model→prompt→compare→bot→app→api). For people who bounce between Claude-for-code and GPT-for-prose and Gemini-for-summary, one $19.99 beats three $20 subs; for single-model loyalists, native app wins on limits/memory/features.

Bluedot is a privacy-first, “invisible” AI notetaker and meeting intelligence platform. Instead of sending a bot into the call as a visible participant, it captures meetings from the user’s own device — via Chrome extension, macOS/Windows desktop app, iOS/Android phone or tablet, and even Apple Watch. That makes it especially attractive for sales calls, recruiter screens, board meetings, client check-ins, and in-person interviews where a recording bot would change the tone of the conversation.

Creatify AI​ is an AI-native video-ad platform built for e-commerce brands, DTC sellers, agencies, and performance marketers. You paste a Shopify, Amazon, or app-product URL (or upload an image / brief), and Creatify scrapes the product, writes ad scripts, picks a realistic AI actor or UGC avatar, adds voiceover, music, and captions, then renders multiple 9:16 / 1:1 / 16:9 ad variants ready to test on Meta, TikTok, YouTube, Snap, and AppLovin. Unlike faceless-shorts tools (InReels) or localization/avatar platforms (HeyGen), Creatify is obsessed with the paid-media loop: generate → clone winning ads → track competitors → launch → read ROAS/CTR → regenerate. Its differentiators are URL-to-video, batch variation (up to ~50 at once), Ad Cloner, Competitor Ad Tracker (10M+ Meta ads), Performance Agent, Ad Launcher, MCP for Claude/ChatGPT, and a credit-metered API for programmatic ad factories.

InReels AI​

#254 New

Video Creation & Editing

InReels AI​ is an all-in-one, vertical-first AI content platform for short-form video. You describe an idea, paste a product URL, or upload a product photo, and it generates a finished Reel/TikTok/YouTube Short-style video with AI script, voiceover, captions, music, avatars, and effects — usually in 2–3 minutes. It covers two big use cases at once: organic faceless content​ (Reddit stories, horror, brainrot, educational shorts) and AI UGC / product video ads​ (avatar spokespersons, product-in-hand compositing, cinematic product ads). Unlike pure video-model playgrounds such as Pika or Runway, InReels is positioned as a “content pipeline,” not just a clip generator: higher-tier plans add automation series, scheduled/auto-posted content, and commercial usage. The pitch is simple: solopreneurs, dropshippers, and small DTC brands can replace five tools (image gen, script writer, voiceover, editor, scheduler) with one affordable workspace.

KreadoAI

#252 New

Video Creation & Editing

KreadoAI​ is an all-in-one AIGC content platform built for global / cross-border marketing, e-commerce, training, and multilingual video production. Instead of hiring voiceover artists, spokespersons, translators, and video editors, a team can log into KreadoAI, paste a product URL / PPT / PDF / script / image, pick one of 1,000+ AI avatars​ and 40,000+ voices across 140+ languages, and generate a finished spokesperson video in minutes. Its core strength is “write once, localize everywhere.”​ A single English script can become a Mandarin spokesperson ad, a Spanish TikTok reel, an Arabic product demo, and a Japanese e-learning clip — with voice cloning, avatar cloning, and lip-synced video translation​ keeping the brand voice consistent across markets. Beyond avatar videos, KreadoAI also covers talking-photo videos, AI fashion/models for e-commerce, digital-human livestream shopping hosts, AI scripts, AI images, background removal, face swap, subtitle/watermark removal, and multi-size ad creatives, making it closer to a multilingual commerce-video studio than a simple text-to-video tool. Pricing runs on a K-coin system: Free gives a small trial pack, Premium starts around $19/mo, Pro around $40/mo​ with API + voice/avatar clones, and Enterprise adds private deployment and SLAs. Reviewers generally praise its language coverage, URL-to-video speed, and e-commerce integrations​ (Shopify / SHOPLINE / Shoplazza-style stores), while critics point to opaque credit math, templated gestures, and weaker lip-sync in rare languages​ versus HeyGen / Synthesia.

Syllaby

#250 New

Video Creation & Editing

Syllaby​ (https://syllaby.io) is an end-to-end “faceless content operating system” for creators, coaches, realtors, local businesses, and small agencies who know they should post on TikTok, Instagram, YouTube, and LinkedIn — but don’t want to be on camera, don’t know what to say, and don’t have an editor. Instead of starting from a finished video like Opus Clip or Vidyo.ai, Syllaby starts from intent: it researches what your audience is searching (questions, keywords, trends, CPC signals), turns those into hook-driven scripts and blog/linkedin/caption drafts, then generates faceless short videos​ using stock/B-roll scenes, AI avatars / Digital Twin presenters, and cloned voices. From there it behaves like a mini social media department — auto-captions, AI thumbnails, a content calendar, and bulk scheduling to TikTok, Reels, Shorts, LinkedIn, Facebook, and YouTube. Under the hood it plugs into modern generative models (Sora 2, Google Veo 3, Seedance, etc. depending on plan), offers voice cloning and custom avatars on mid/high tiers, ships iOS/Android apps, and has begun exposing an API + MCP server for Claude/Cursor workflows. Pricing starts at 9/mo(Starter)∗∗andscalesto∗∗29 Basic / 78Standard/153 Premium, with credits consumed per video/script/idea rather than flat unlimited — which is why reviewers love its pipeline (“idea → script → faceless video → scheduled post in one login”) but warn that the credit math gets expensive at agency volume. In short: Opus Clip repurposes what you already recorded, HeyGen perfects avatars, Pictory converts text into video — Syllaby sits earlier in the funnel and answers “what should I even post, and can I ship it without showing my face.”

Turns podcasts, webinars, YouTube videos, and interviews into captioned short clips, then schedules/publishes them across TikTok, Reels, Shorts, LinkedIn, Facebook, X, and Pinterest from one dashboard.

​ Munch is a strong “long video → social clips” tool, especially for podcasts, webinars, and interviews. Rate it ~3.5–4.6/5 depending on source: great time-saver, but priced for teams, not casual creators, and weaker on deep editing + billing experience.

AdCreative.ai is an AI ad creative platform founded in 2021 and acquired by Appier in 2025. It is not a general-purpose design tool​ (distinct from Canva or Firefly), but an ad creative production engine purpose-built for performance advertisers, DTC brands, and agencies. Its core selling point is "generated creatives come with a conversion prediction score"​ — the model is trained on 1B+ ad creatives​ generated on the platform combined with real-world ad performance data.

Adobe Firefly — Full Introduction (2026, Adobe Inc.) Adobe Firefly is Adobe's commercially-safe generative AI suite​ for images, video, vectors, and audio, launched in beta March 2023 and now sitting across the Creative Cloud ecosystem. Its defining position is training data: Firefly's own models are trained exclusively on licensed Adobe Stock, public-domain works, and openly…

Ahrefs — Full Introduction (2026, Ahrefs Inc. / Singapore + Palo Alto) 1. One-line positioning Founded in 2011 and operating out of Singapore and Palo Alto, Ahrefs was created by Dmytro Gerasymenko, while Tim Soulo acts as the platform’s primary public‑facing representative. Long‑established as the incumbent backlink‑index tool, it has since expanded into an end‑to‑end…

aider (Paul-Gauthier Aider-AI Apache-2.0 Apr-2023 active-2026-09 not-shut-down no-acquisition 48k-stars 6.8M-installs 15B-tok-wk v0.86.2-Feb-2026 maintenance-slowed-not-archived MCP-roadmap-unshipped) = terminal-native-git-aware-AI-pair-programmer BYOK-not-autonomous-not-IDE-plugin: chat-SEARCH-REPLACE-apply-lint-test-auto-commit /undo-reverts-commit repo-map-tree-sitter architect-mode-planner+editor-split /add-explicit /web-url-to-markdown image-voice-whisper 100plus-LLMs-Claude-Sonnet-Opus4 GPT-4o-o1-o3mini-GPT-5 DeepSeek-R1-V3-V4Flash Gemini-1.5-2.5Pro Qwen Ollama-LM-Studio-LiteLLM vim-emacs-Helix-Zed-VSCode-JB-sidepane SSH-tmux-Docker no-tab-complete no-GUI no-MCP no-browser-tool; CLI $0 Apache-2.0 LLM-BYO $0-local-or-5-50-mo typical no-paid-tier; category ai-productivity-data pipeline_stage none; competitors Claude-Code OpenAI-Codex-CLI Gemini-CLI-Qwen-Code Continue Cline-Roo Cursor Windsurf GitHub-Copilot JetBrains-AI Devin Factory; not-ElevenLabs-AIVA-PixVerse-Synthesia-Voicemod-overlap.

Airtable AI = generative/agentic AI layer inside Airtable's no-code relational database

AIVA is "the AI composer that hands back an editable score, not just a rendered track." No lyrics, no vocals: you choose from 250+ style presets (Modern Cinematic, Symphonic, Chinese, Jazz, Ambient, Electronic, Tango…) or upload a reference MIDI/audio to train a custom style model, set key/tempo/length/emotion, and it composes a full multi-section arrangement. The built-in editor shows piano-roll score — move a note, change velocity, re-voice the brass, swap instruments — then export MIDI to finish in Logic/Cubase/Pro Tools, or WAV+STEMS on Pro for direct-to-media. 2017 it became the first AI registered as a composer by SACEM, which is why its copyright ladder (Free→Standard→Pro) reads like a music-publishing contract, not a SaaS EULA. Pitch: Suno writes the sung single; AIVA drafts the film/game score you own and can re-orchestrate by hand.

Akool

#7

Akool = Palo Alto enterprise AI video suite (founded 2021/2022 by Dr. Jiajun "Jeff" Lu, ex-Apple Face ID / Google Cloud video, #1 on Inc. 5000 2025, 10M+ users / 73K+ companies / 300M+ assets, Coca-Cola + Qatar Airways + Canon + Logitech + AWS + Google Cloud as customers) that bundles AI avatar video, real-time streaming avatars, face swap up to 16K, video translation in 155+ languages with diffusion lip-sync, talking photo, voice clone, text/image/reference-to-video, PPT-to-video, and holographic event display — all behind one pooled credit meter and a REST API (x-api-key). Free 720p/5-min watermark; Pro $21/mo annual ($30 monthly, 600 cr); Pro Max $79/mo annual ($119 monthly, 1,200 cr, API); Business $350/mo annual ($500 monthly, 6,000 cr, Studio Avatar, commercial license); Enterprise custom with non-expiring credits. It is the breadth + face-swap + live-interaction bench — distinct from HeyGen (script-to-presenter realism, 175 langs, SCORM), Synthesia (L&D-only), and Runway/Kling/Veo (pure footage, no avatar product).

Amazon Q Developer is in sunset — not shut down yet, but AWS closed new signups 2026-05-15 and will end support for IDE plugins + paid subscriptions on 2027-04-30 (successor = Kiro).​ Per your rule ("if closed, warn and do not output intro"), I warn: not outputting full English introduction body, but giving the rest in table form with status flag upfront.

Anyword = an LLM-agnostic conversion-prediction layer for marketing copy: generates ad headlines / email subjects / LP sections / SMS / product blurbs / blog drafts, then scores every variant 0–100 on predicted CTR+conversion with demographic + emotion panels

Headquartered in Prague, Czech Republic, Apify Technologies s.r.o. was founded in 2015 by Jan Curn and Jakub Balada, participants of the 2015/2016 YC Fellowship program. The company operates with a VC-independent leaning while retaining formal VC backing. Prior to its first institutional financing, Apify achieved profitability through a bootstrapped operation model. It secured a €2.8M funding round in April 2024 led by J&T Ventures and Reflex Capital, with total capital raised reaching approximately $4.5M per Simplify data. As a YC alumni with limited venture funding, Apify remains unacquired, non-public, non-nonprofit, and is not classified as a Big Tech platform primitive. It falls into the VC-independent vendor bucket alongside n8n, Databox, Typeform and Documenso, clearly separated from pure bootstrapped players including Todoist and QuickChart. Apify provides a stable, official first-party GA MCP server hosted at mcp.apify.com, adopting OAuth streamable-HTTP transport while discontinuing SSE support as of April 2026. Its official open-source MIT-licensed MCP server @apify/actors-mcp-server holds 3.7K to 4.8K GitHub stars, unlocking a massive tool ecosystem of 26K–30K Apify Actors available for agent invocation. Ecosystem-wise, Apify joins the top-tier stable first-party GA server cluster with Klaviyo, Feedly, Slack, Databox and Tavily. It is differentiated from n8n dual-GA architecture, Google Sheets/Gmail first-party preview stack, Typeform beta-grade MCP, Nanonets Composio-exclusive integration model, and the zero-first-party MCP cohort covering Documenso, QuickChart, Semantic Scholar and Zotero.

Auphonic (Auphonic GmbH, Austria 2012/2013) Status: Active (founded 2012/2013 by Hans-Jürgen Schüler et al., Graz; 2 hrs free/mo permanent; 2026-01 RMS-based loudness for Audible/ACX, 2025-12 Denoising Editor; auphonic.com + API + CLI live 2026-09). Category ai-productivity-data; pipeline_stage=none. Not a podcast editor (→ Descript), not a live call denoiser (→ Krisp)​ — it is the broadcast-standard…

Bannerbear — Template-Based Image/Video/PDF Generation API with AI Layers (BG Removal/Face Detect/AI BG Gen), Workflows, Webhooks, 50+ No-Code Integrations & Official MCP Server (Bootstrap Bali-2018/2020, Active) Bannerbear (sole proprietor / single-member bootstrap, Jon Yongfook Cockle). Origin: Previewmojo (OG-image generator) Nov 2019 → rebranded Bannerbear Jan 2020; REST API + template editor Mar 2020. Legal/base Bali, Indonesia; founder Singapore/Bali. Unfunded, no acquisition, no shutdown.​ ~$1M ARR Sep 2025, 7 staff Jun 2026.

Beautiful.ai (SF, 2016, Mitch Grasso, ~$45M) = Smart Slides auto-layout engine + DesignerBot/Create-with-AI prompt draft, the brand-governance lane between PowerPoint (manual) and Gamma (pure generative). Your 28th CPT card, opens generate-presentations​ stage (16-stage enum now). Below is the table-first version.

Bolt.new = StackBlitz's browser-native AI app builder that turns a natural-language prompt into a running full-stack web app (React/Next/Vue/Svelte/Vite/Express + Tailwind/Shadcn/MUI) by generating the files, npm-installing, booting a dev server, and live-previewing entirely inside one browser tab via WebContainers — then one-click deploying to Bolt Cloud/Netlify with built-in databases, auth, custom domains, SEO. Token-metered: Free 300K/day + 1M/mo (watermark, no rollover), Pro $25/mo ($18 ann) 10M/mo no daily cap + rollover 1 cycle, Teams $30/seat/mo ($27 ann) same 10M/seat + centralized billing/private NPM/design-system prompts, Enterprise custom (SSO/SAML/audit/BYOK). Agentic loop: on build error the agent reads the stack and self-heals before you ask. It is the "prompt→running app" bench, not a media tool (Veo/HeyGen), not a DB AI layer (Airtable AI), not a text-repurpose tool (Lumen5/Pictory/Opus).

Braintrust (Braintrust Data, Inc.) is active (not shut down).​ SF, founded 2023 by Ankur Goyal (ex-Impira), Series B 80MledbyICONIQFeb2026( 800M val, $124M total). AI-native evals + observability for LLM apps/agents. No acquisition, no sunset. Category ai-productivity-data, pipeline_stage=none. File under Dev​ (LLM-eval/observability sublane). This is a re-run of your prior Braintrust turn with verified 2026-09 figures.

Buffer (Boston, 2010, Joel Gascoigne) = per-channel-priced social scheduler + unlimited AI Assistant + light community inbox/analytics, the solopreneur lane between manual posting and enterprise suites (Hootsuite/Sprout).  Snapshot

Canva Magic Studio (Canva Pty Ltd, Sydney 2012–2013, 190M+ MAU, 2026 "Creative OS" bundling Affinity + Canva Sheets + Canva Code) = the in-editor AI layer over Canva's drag-drop design surface​ — Magic Write / Magic Design / Dream Lab / Magic Edit-Expand-Grab-Morph-Animate-Switch / Magic Charts-Insights / Conversational AI / Canva Code — not a standalone SKU, not Midjourney, not Figma.