Tap a chip above to filter. Cards group by category in the default view. Each card shows AI/Agentic Level and MCP support inline.
S
Chat
General-purpose AI assistant + agentic Cowork mode
Anthropic's general AI assistant.
AgenticVery High
MCPNative
Strong at writing, reasoning, and code. Long context window. Cowork mode runs agentic workflows on your desktop / mobile — the differentiator most other chat apps don't have.
Free tier · Pro $20/mo · Max from $100/mo
Best for: any captain who can only afford one AI subscription and needs writing + reasoning + agentic capability all in one.
Best default for most men. If you only buy one, this is the one.
A
Chat
General-purpose AI assistant + computer-use Agent
OpenAI's assistant. Largest install base.
AgenticVery High
MCPYes
GPT-5 family models, image generation built into the chat, custom GPTs marketplace, voice mode. The default for many guys who tried AI first. ChatGPT Agent (formerly Operator) handles computer-use tasks.
Free tier · Plus $20/mo · Pro $200/mo · Business / Enterprise scales up
Redundancy check — Heavily overlaps Claude for general writing & reasoning. Pick one as primary; second is optional cross-check.
Best for: captains who want strong image generation + voice mode + a sanity-check second opinion alongside Claude.
Plus at $20 is the right tier as a #2 for image gen + sanity check. Pro at $200/mo is overkill for a family-captain stack — nobody in this cohort needs that ceiling.
A
Chat
Google Workspace-native AI assistant
Google's AI, baked into Workspace.
AgenticHigh
MCPNo
Strong free tier and the longest context window of the big three. Best when you live in Google Drive, Docs, Gmail — it sees your stuff with permission.
Free tier · AI Pro $19.99/mo (often included in Google One)
Redundancy check — Overlaps Claude / ChatGPT. Becomes essential only if your work is Google-native.
Best for: captains whose ministry/business runs in Google Drive, Docs, Sheets, Gmail.
Skip unless you're deep in Google. Free tier is decent for the curious.
B
Chat
X-integrated AI chat with live X data
xAI's chat, built into X (Twitter).
AgenticMedium
MCPNo
Best for live X data and an unfiltered conversational style. Limited utility outside that lane for most captains.
Bundled with X Premium $8/mo+ · Standalone Super Grok ~$30/mo
Redundancy check — Heavily redundant with ChatGPT / Claude for general work.
Best for: captains who post on X heavily and want real-time X-data context for replies and posts.
Skip unless you're a heavy X user or want a contrarian second take. Don't pay for it as a primary.
A
Chat
Search-first AI with citations
Search-first AI. Citations built in.
AgenticMedium
MCPNo
Real-time web search wrapped in chat with proper citations. Replaces "Google for research" for most captains. Pro tier now bundles Comet, the agentic AI browser.
Free tier · Pro $20/mo (includes Comet browser)
Redundancy check — Overlaps ChatGPT / Gemini search modes. Standalone is cleaner if research is your bottleneck.
Best for: captains doing research-heavy work — sermon prep, due diligence, fact-checking — who need cited sources.
Free tier is enough for 90%. Pro only if you research daily for work.
A
Chat
Microsoft 365 AI assistant
Microsoft's AI in Edge, Windows, M365.
AgenticMedium
MCPYes
Integrated across Windows, Edge, and Microsoft 365 apps. The right pick if you live in Outlook / Word / Excel and your IT shop is Microsoft. Microsoft adopted Model Context Protocol (MCP) in late 2025.
Free tier · Copilot Pro $20/mo · M365 Copilot $30/mo per user
Redundancy check — Redundant with Claude / ChatGPT for chat. Earned its keep when paired with M365.
Best for: captains whose business runs on Word, Excel, Outlook, and Teams.
Pay only if your daily work is in Word / Excel / Outlook. Otherwise skip.
C
Chat
Open-weight AI assistant (free, capable)
Open-weight Chinese model. Free. Capable.
AgenticMedium
MCPNo
DeepSeek-V3 and reasoning models like R1 often match GPT-4-class quality at zero cost. Free web app + cheap API. Worth knowing for cost-sensitive captains and as a sanity-check second opinion.
Free (web app) · API ~$0.27 per 1M input tokens (roughly 1/30th of GPT-4's price)
Redundancy check — Overlaps Claude / ChatGPT for general work. Free tier is the killer feature.
Best for: captains on a $0 budget who still want frontier-class output. Sanity-check tab next to your primary AI.
Worth a sanity-check tab next to Claude. Don't make it primary — but free firepower is real.
S
Voice
Speech-to-text model (open source)
Industry-leading transcription.
AgenticLow
MCPN/A
Open source model with leading accuracy across accents and noise. Available via OpenAI API ($0.006/min) or local apps (MacWhisper, etc.). Most other transcription tools use Whisper under the hood.
Free (local) · ~$0.006/min via OpenAI API
Redundancy check — If your iPhone already transcribes Voice Memos to your taste, you may not need Whisper.
Best for: captains who record voice memos / interviews / sermon dictation and want clean transcripts.
Run locally on a Mac if you have one. Otherwise let Apple or Otter handle it.
S
Voice
Built-in iPhone voice recording + transcription
Built-in iPhone transcription.
AgenticLow
MCPNo
Voice Memos transcribes your recordings on-device on iOS 18+ with Apple Intelligence. Apple Notes records audio with live transcripts too. May be all you need.
Free (with iOS 18+ on a recent iPhone)
Redundancy check — Replaces Otter and most paid transcription tools for personal use.
Best for: captains who already have an iPhone and want zero-friction voice capture.
Try this first. If it covers your captures, you're done.
A
Voice
Live meeting transcription service
Live meeting transcription, speaker ID.
AgenticMedium
MCPNo
Joins Zoom / Google Meet / Teams calls and transcribes in real time with speakers labeled. Searchable archive of meetings is the killer feature.
Free tier (300 min/mo) · Pro $16.99/mo · Business $30/mo per user
Redundancy check — If your meeting platform already records with transcripts (Zoom AI Companion, Google Meet Gemini), you may not need Otter.
Best for: captains with 3+ meetings/week they wish they'd captured (counseling, coaching calls, board meetings).
Worth it if you have 3+ meetings/week you wish you'd captured. Otherwise skip.
A
Voice
Audio/video editor that edits transcripts
Edit audio & video by editing the transcript.
AgenticMedium
MCPNo
Type to delete words, type to overdub, AI fills in gaps. Strong for podcasters and short-video creators. Word-level editing of audio is genuinely magic the first time.
Free tier · Hobbyist $19/mo · Creator $35/mo
Redundancy check — Overlaps with traditional editors (Premiere, Final Cut) for video; the transcript-edit workflow is unique.
Best for: captains starting a podcast, YouTube channel, or audio-content creation.
Buy it the month you start a podcast or YouTube channel. Skip otherwise.
A
Voice
One-click audio cleanup tool
Fixes terrible audio in one click. Free.
AgenticLow
MCPNo
Drop in a noisy phone-mic recording — sermon prep, voice memo, half-blown-out interview — get back broadcast-quality audio. The fastest 'this is magic' moment in audio AI. Free for now as Adobe pushes it as a Creative Cloud upsell.
Free (web app) · Unlimited inside Premiere Pro for CC subscribers
Redundancy check — Different shape from Whisper (Whisper transcribes; Enhance cleans audio for listening).
Best for: pastors, podcasters, and ministry leaders recording on phones in less-than-ideal rooms.
First tool I'd hand a pastor recording on his iPhone. Try it once and you'll never not use it.
B
Image
AI image generation (artistic / stylized)
Best aesthetic quality for stylized images.
AgenticLow
MCPNo
Web app or Discord-based image generation. Strongest house style of any image model — gets you to publishable visuals fastest if your taste matches its aesthetic.
Basic $10/mo · Standard $30/mo · Pro $60/mo
Redundancy check — Replaces DALL-E / Firefly for most creative use. Less integrated with workflows.
Best for: captains making weekly visual content (sermon graphics, social posts, blog hero images).
Worth it if you make weekly visual content. Skip if image gen is occasional — ChatGPT's built-in is enough.
A
Image
Image gen baked into ChatGPT
Image gen baked into ChatGPT.
AgenticLow
MCPYes
GPT-Image-1 / DALL-E inside ChatGPT. Convenient because it's right where you're already chatting. Quality is good, not best-in-class.
Included with ChatGPT Plus $20/mo
Redundancy check — Redundant with Midjourney / Firefly if you have either.
Best for: captains who already pay for ChatGPT and only need occasional image generation in-flow.
The right pick if you only generate the occasional image. Don't pay for ChatGPT just for this.
B
Image
Commercially-safe AI image generation
Commercially safe, trained on licensed content.
AgenticLow
MCPNo
Firefly is trained on Adobe Stock + licensed content, so output is safer for commercial use. Lives inside Photoshop / Illustrator if you have Creative Cloud.
Standalone from $4.99/mo · Included with Creative Cloud subscriptions
Redundancy check — Pricing-wise hard to beat if you already pay for CC. Standalone is duplicate of Midjourney.
Best for: captains creating commercial work (logos, products to sell, branded marketing) where IP-safety matters.
Use it ONLY as part of a Creative Cloud bundle. The $4.99 standalone is a trap — captains who buy it will then need Photoshop or Illustrator to actually USE what Firefly generates, and that's the real cost. If you don't already pay for CC, pick Midjourney or Ideogram instead.
C
Image
AI image generation specialized for text-in-image
Best for text-in-image (logos, posters).
AgenticLow
MCPNo
The image model that actually spells words right. Strong for logos, posters, social cards with copy on them — the place other image models fall apart.
Free tier (limited) · Plus $7/mo · Pro $20/mo
Redundancy check — Specialty tool. Doesn't replace a general image model.
Best for: captains making flyers, sermon-graphic cards, posters — anything with text rendered IN the image.
$7 stack-add for the captain who makes flyers, sermon graphics, or marketing posts.
B
Image
Open-source image generation models
Open source. Free if you self-host.
AgenticLow
MCPN/A
Stable Diffusion is the original open foundation model behind many image apps. FLUX.1 (Black Forest Labs, 2024+) has eclipsed it for quality and is now the open-weight champion. Run either locally (free, requires a decent GPU) or via paid hosted services.
Free (self-hosted) · Hosted services from $0–$30/mo · FLUX.1 [pro] from $0.05/image via API
Redundancy check — Niche unless you need full control or local privacy. FLUX.1 is technically separate from Stable Diffusion but lives in the same captain category.
Best for: technical captains who want to own the image-gen stack (local privacy, custom fine-tuning, no per-image cost).
Skip unless you're technical and want to own the stack. Most captains are better served by Midjourney or Firefly.
C
Image
AI image upscaler / enhancer
AI upscaler / enhancer. Print-quality output.
AgenticLow
MCPNo
Takes a small or low-quality image and makes it print-ready. Strong for sermon graphics, marketing flyers, merch — anything that needs to look good on paper or scaled up to a banner.
Pro $39/mo · Premium $99/mo · no free tier (a 7-day trial sometimes)
Redundancy check — Different from Midjourney / Firefly — those generate, Magnific enhances. Pair, don't replace.
Best for: captains making print-quality flyers, posters, merch — output that needs to look sharp at large sizes.
Skip for most captains. Midjourney's built-in upscale + Ideogram already cover the family-captain print use case. Magnific is $39/mo with no free tier — only worth it if you're running print-shop volumes weekly.
A
Notes
AI inside the Notion workspace
AI inside the Notion docs you already use.
AgenticMedium
MCPNo
Q&A across your workspace, draft / summarize / translate inline, AI databases. Best if Notion is already your second brain.
Notion Free / Plus $10 + AI add-on $10/mo · Business $24+/mo includes AI
Redundancy check — Overlaps Apple Notes + Apple Intelligence and Obsidian + plugins for personal use.
Best for: captains running their business / family ops in Notion who want AI Q&A across their docs.
Add the AI on if Notion is your daily driver. Skip if you only open Notion monthly.
A
Notes
Local-first knowledge graph w/ AI plugins
Local-first knowledge graph.
AgenticMedium
MCPPlugin
Markdown files on your hard drive, linked into a graph. Pair with Smart Connections / Copilot plugin for AI search across your notes — without sending them to the cloud.
Obsidian free for personal · Some plugins free, some paid
Redundancy check — Replaces Notion for the privacy-conscious captain. Different model than database apps.
Best for: captains who want files on disk (portable, owned, private) and AI search over their own corpus.
For the captain who wants files on disk, owned, portable. Curve is steeper than Notion's.
A
Notes
Built-in iOS/macOS notes w/ Apple Intelligence
Built-in. Now AI-aware.
AgenticLow
MCPNo
With Apple Intelligence: summarize notes, audio recordings with transcripts, writing tools. Already on your phone, already syncs to your Mac.
Free (with iOS / macOS 18+ on a recent device)
Redundancy check — Replaces a lot of paid note tools for personal use.
Best for: captains on Apple devices who want zero-friction note-taking with AI baked in.
Try this first. Move to Notion / Obsidian only if you hit a wall.
A
Notes
AI meeting notes (Mac-only)
AI meeting notes that listen to your audio context.
AgenticMedium
MCPNo
Mac-only. Listens to your Zoom / Meet / Teams audio, generates structured meeting notes. Different from Otter — less raw transcript, more clean notes.
Free tier · Pro $18/mo · Business $35/mo per user
Redundancy check — Otter.ai is the live alternative. Granola feels less like a recorder, more like a notetaker.
Best for: Mac-using captains with back-to-back calls who want clean structured notes without active note-taking.
Worth it if your day is back-to-back calls. Free tier is enough to test.
A
Notes
AI research notebook (Google) — free
Google's AI research notebook. Free.
AgenticMedium
MCPNo
Upload PDFs, docs, websites — NotebookLM ingests them, lets you Q&A across the corpus, and generates podcast-style audio summaries that sound shockingly natural. Killer for digesting books, sermon prep, or research-heavy weeks.
Free (Google account)
Redundancy check — Different shape from Notion AI — NotebookLM is research-focused; Notion is workspace.
Best for: captains digesting books, study materials, or large document sets — especially for sermon prep.
Drop your bootcamp materials in, ask 'what's the captain's role in Week 3?' Pure leverage. Free.
S
Automation
Desktop / mobile AI agent built into Claude
Desktop / mobile agent built into Claude.
AgenticHighest
MCPNative
Run files, automate browser tasks, execute multi-step workflows on your machine with the captain still on the bridge. The bootcamp's primary tool.
Included with Claude Pro / Max (Pro $20/mo, Max from $100/mo)
Redundancy check — Replaces basic Zapier for personal workflows. Doesn't replace Zapier for multi-app SaaS chains.
Best for: captains who want a true AI adjutant on their machine — files, browser, code, all under captain oversight.
This is the differentiator. The reason Claude is the bootcamp's default.
A
Automation
No-code automation between apps
No-code automation between apps.
AgenticMedium
MCPPlugin
The granddaddy of automation. 7,000+ app integrations. AI features now baked across the product. Free tier handles light use.
Free tier (5 single-step Zaps) · Paid from $19.99/mo
Redundancy check — Overlaps Make.com (Make is more flexible / cheaper at scale).
Best for: captains who need to wire up SaaS apps that don't talk to each other (Mailchimp → Google Sheets, etc.).
Free tier is fine for the captain with 1–2 simple automations. Pay only if you have a real workflow.
B
Automation
Visual automation platform
More flexible than Zapier. Visual.
AgenticMedium
MCPPlugin
Branching, looping, complex visual workflows. Better unit economics than Zapier above the free tier. Steeper learning curve.
Free tier (1,000 ops/mo) · Paid from $9/mo
Redundancy check — Functional overlap with Zapier — pick one.
Best for: captains whose automations branch / loop / hit the data-volume cliff in Zapier.
Pick this over Zapier if your automations branch or loop. Otherwise Zapier's marketplace is bigger.
B
Automation
Open-source automation, self-hostable
Open source automation. Self-hosted option.
AgenticHigh
MCPNative
Same shape as Zapier / Make but you can self-host for full control. Strong AI nodes, native Model Context Protocol (MCP) support added in 2025. The captain's choice for owning the stack.
Self-host free · Cloud from $20/mo · Pro $50/mo
Redundancy check — Replaces Zapier / Make for the technical captain.
Best for: technical captains who want full control + data sovereignty + zero per-execution fees.
Skip unless you're already running a homelab or care about data sovereignty.
B
Automation
AI agent automation — Zapier with reasoning
AI agent automation. Zapier with reasoning.
AgenticVery High
MCPNo
Build agents that read email, summarize meetings, draft replies, manage CRMs — with AI reasoning in the loop, not just deterministic if-this-then-that. Strong for solo founders who can't afford an assistant but can afford $50/mo.
Free tier (limited) · Pro from $49.99/mo · Business from $199.99/mo
Redundancy check — Overlaps Zapier-with-AI-nodes. Lindy is more agentic; Zapier is more deterministic.
Best for: captains running real businesses with email overload who can't afford a virtual assistant.
For captains running real businesses with email overload. Skip if Zapier+AI covers you.
S
Coding
Terminal coding agent (Anthropic)
Anthropic's terminal coding agent.
AgenticHighest
MCPNative
Run from your terminal. Reads your repo, writes / edits code, runs tests, opens PRs — under your supervision. Feels like a senior engineer pair-programming.
Included with Claude Pro / Max
Redundancy check — Different shape than Cursor (terminal vs editor). Many devs run both.
Best for: captains shipping real code who live in a terminal and want autonomous-with-oversight coding.
For captains shipping real code — this is the bootcamp's recommendation.
A
Coding
AI-first code editor (VS Code fork)
AI-first code editor. Fork of VS Code.
AgenticHigh
MCPYes
Tab autocomplete that's eerily good. Composer / Agent for multi-file edits. Industry standard for AI-assisted coding in 2025–26.
Free tier (limited) · Pro $20/mo · Business $40/mo per user
Redundancy check — Different shape than Claude Code. Comparable to Windsurf.
Best for: captains who code in a graphical editor and want AI assistance baked into every keystroke.
Pick this if you live in a graphical editor; Claude Code if you live in a terminal.
B
Coding
AI coding assistant in GitHub / VS Code
Tight GitHub / VS Code integration.
AgenticHigh
MCPYes
Autocomplete, chat, PR summaries inside GitHub and VS Code. Now multi-model (Claude, GPT, Gemini available). Free tier for individuals.
Free tier (limited) · Pro $10/mo · Business $19/mo per user
Redundancy check — Cursor / Windsurf are stronger AI editors; Copilot wins on GitHub integration.
Best for: captains on GitHub-native teams who want AI tightly bound to their code-review and PR flow.
For developers already on GitHub-native teams, $10/mo is fair. For non-developers learning to ship, Cursor or Windsurf at the same price are stronger picks — Copilot's autocomplete shines only when you can already read code.
B
Coding
AI code editor — Cursor competitor
Cursor competitor. Strong agent mode.
AgenticHigh
MCPYes
Codeium's AI editor — same idea as Cursor with a different feel. Strong agent / cascade mode. Free tier is generous.
Free tier · Pro from $15/mo
Redundancy check — Direct competitor to Cursor. Pick one.
Best for: captains evaluating AI editors — try both Cursor and Windsurf, keep whichever clicks.
Worth a 1-week trial alongside Cursor; whichever feels more natural is the winner.
B
Coding
Open-source AI pair programmer (terminal)
Open-source AI pair programmer. Model-agnostic.
AgenticHigh
MCPYes
Terminal-based, language-agnostic, works with any model API (Claude, GPT, Gemini, Llama). Same shape as Claude Code but you bring your own model. Pythonic, scriptable, MIT licensed.
Free (open source) · You pay only for the model API calls you make
Redundancy check — Different shape from Cursor (terminal vs editor). Closest to Claude Code, but supports any model.
Best for: technical captains who want full control over which model they use + zero subscription lock-in.
For technical captains who want full control + multi-model flexibility. Skip if Claude Code already works.
B
Video
AI voice generation / cloning
Industry-leading AI voice / TTS.
AgenticLow
MCPNo
Voice cloning, narration, multilingual dubbing. The voices on most polished AI-narrated content come from here.
Free tier (limited) · Starter $5/mo · Creator $22/mo · Pro $99/mo
Redundancy check — Different shape than Whisper (TTS vs STT).
Best for: captains producing audiobooks, narrated devotionals, or polished video voiceovers at scale.
For the captain producing audiobooks, narrated devotionals, or polished video voiceovers.
C
Video
AI talking-head avatar video
AI avatars. Talking-head video without filming.
AgenticLow
MCPNo
Generate talking-head video from a script and a photo / clone of yourself. Real-time avatars in some plans. Strong for course / training content at scale.
Free tier (1 min) · Creator $24/mo · Team $89/mo per user
Redundancy check — Niche — replaces filming, not editing.
Skip for personal cohort use (real face matters). Worth it for high-volume course / sales content.
Confirmed skip for a captain cohort. Your real face matters to your men. HeyGen earns its keep ONLY when you're producing course or sales content at a volume you can't physically film yourself.
B
Video
Long-video to short-form auto-clipper
Auto-clip long videos into shorts.
AgenticMedium
MCPNo
Drop in a long video, get back ~10 short-form clips with captions, cropping, hooks. Saves hours per upload if you publish to TikTok / Reels / Shorts.
Free tier (60 min/mo) · Starter $9/mo · Pro $29/mo
Redundancy check — Descript can do similar with more manual work.
Best for: captains publishing weekly to TikTok / Instagram Reels / YouTube Shorts.
Worth $9 if you publish weekly to short-form social. Skip otherwise.
B
Video
AI music generation
AI music generation. Hum a vibe, get a song.
AgenticLow
MCPNo
Type 'upbeat acoustic guitar intro for a men's ministry podcast, 30 seconds' — get a custom track in 60 seconds. Quality is shockingly good for podcast intros, sermon backgrounds, and video soundtracks.
Free tier (50 credits/day) · Pro $10/mo · Premier $30/mo
Redundancy check — Different shape from ElevenLabs (Suno = music; ElevenLabs = voice).
Best for: captains making audio/video content who need custom intros, beds, or full tracks without licensing royalty-free music.
For captains making any audio/video content. Don't sleep on the free tier — it's enough for casual use.
B
Agentic Frontier
Browser-based full-stack coding agent
Build full apps from a prompt — no laptop required.
AgenticVery High
MCPYes
Browser-based. Type what you want; Replit Agent codes, deploys, and iterates. Zero local setup. Strong for captains who want to ship without learning a terminal.
Free tier (limited) · Core $20/mo · Teams $35/mo per user
Redundancy check — Different shape than Claude Code (browser vs terminal). Lovable / Bolt are direct competitors.
Best for: non-technical captains who want to ship a working web app or simple SaaS without installing anything.
Best browser-based agentic builder for captains who don't want to touch a terminal.
B
Agentic Frontier
Text-to-app generator (full-stack)
Ship a full-stack app from a prompt.
AgenticVery High
MCPNo
Lovable generates React + Tailwind apps with backend logic from natural-language prompts. Strong for landing pages, internal tools, simple SaaS MVPs. One of the most-loved tools of 2025–26 for non-developers.
Free tier (50 messages/day) · Pro $20/mo · Teams $30/mo
Redundancy check — Direct competitor to Bolt.new and v0.dev. All three solve similar problems differently.
Best for: captains who want a working web app shipped this weekend without learning React.
First pick for non-coders who want to ship something real. Free tier gets you a working prototype.
B
Agentic Frontier
Text-to-React UI generator (Vercel)
Vercel's text-to-UI generator.
AgenticHigh
MCPNo
Type a UI description; v0 generates React + Tailwind components you can copy into your codebase or deploy via Vercel. UI-focused, less full-stack than Lovable / Bolt.
Free tier · Premium $20/mo · Team $30/mo
Redundancy check — Overlaps Lovable / Bolt for UI generation. Best when you only need the front-end.
Best for: captains who already have a backend and just need polished React UI fast.
Pick this over Lovable if you want UI-only output that drops into an existing codebase.
C
Agentic Frontier
Full-stack agentic builder (StackBlitz)
StackBlitz's in-browser full-stack agent.
AgenticVery High
MCPNo
Build, test, and deploy full-stack web apps in a single browser tab. Faster iteration than most competitors thanks to StackBlitz's WebContainer tech. Strong agentic mode that fixes its own errors.
Free tier · Pro $20/mo · Pro 50 $50/mo
Redundancy check — Direct competitor to Lovable.dev. Both are excellent — Bolt is faster, Lovable has better UX.
Best for: captains shipping prototypes fast — internal tools, MVPs, ministry-facing simple apps.
Bolt and Lovable both work. Try the free tier of each. Whichever lands first, use that one.
B
Agentic Frontier
Autonomous web agent
General-purpose autonomous AI agent.
AgenticHighest
MCPPlugin
Made by Monica AI (China). Manus runs autonomously — research a topic, build a presentation, draft a deck, fill out forms, browse and synthesize. Made significant waves in early-mid 2025 as the first widely-accessible autonomous web agent.
Free tier (300 credits) · Premium ~$39/mo
Redundancy check — Comparable to ChatGPT Agent. Manus is more research / synthesis-focused; ChatGPT Agent is more computer-use.
Best for: captains with research-heavy projects (due diligence, market scans, prep docs) who want autonomous synthesis.
Worth a $39 month-long test if you have a research-heavy quarter. Free tier shows the shape.
B
Agentic Frontier
Agentic AI browser (Perplexity)
Perplexity's AI-powered agentic browser.
AgenticHigh
MCPNo
A full Chrome-replacement browser with Perplexity AI baked in. Comet can browse on your behalf, summarize tabs, fill forms, run multi-step research tasks. Released as a Pro perk through 2025–26.
Included with Perplexity Pro $20/mo
Redundancy check — Overlaps ChatGPT Agent and Dia Browser. Comet is the most search-research-flavored of the three.
Best for: research-heavy captains who already pay for Perplexity Pro and want their browser to be agentic.
Free with Perplexity Pro. If you have Perplexity, you have Comet — try it.
C
Agentic Frontier
AI-native browser (Browser Company)
AI-first browser from the makers of Arc.
AgenticHigh
MCPPlugin
The Browser Company shifted from Arc to Dia in 2025–26. Dia treats AI as a first-class browser feature: every URL bar entry can be a chat, every page can be summarized, every tab is queryable. Beta-pricing through 2026 Q2.
Free during beta (Mac, then Windows) · paid tiers TBD
Redundancy check — Overlaps Comet and ChatGPT Agent for agentic browsing. Dia is more 'AI-native UX' than 'autonomous agent'.
Best for: captains whose primary tool is the browser and who want AI woven through every action there.
Free during beta. Easy to try alongside your current browser without commitment.
B
Agentic Frontier
Computer-use agent (OpenAI)
OpenAI's computer-use agent.
AgenticHighest
MCPYes
Formerly Operator. ChatGPT Agent operates a browser (and increasingly other apps) on your behalf — fills forms, books appointments, places orders, runs research. Bundled with paid ChatGPT plans.
Included with ChatGPT Plus $20/mo (limited) · Pro $200/mo (heavy use)
Redundancy check — Comparable to Manus and Comet. ChatGPT Agent is the most general-purpose 'do this task' agent of the three.
Best for: captains who already pay for ChatGPT Plus / Pro and want a hands-off task agent.
If you have ChatGPT Plus, you have Agent. Worth experimenting once a month with a real task.
C
Agentic Frontier
Open-source desktop agent (Block)
Block's open-source local desktop agent.
AgenticVery High
MCPNative
Open-source, MCP-native, runs on your laptop. Made by Block (Square's parent). Strong choice for captains who want an autonomous agent on their own machine without a SaaS subscription.
Free, open source · You provide your own model API key
Redundancy check — Different shape than Cowork (open-source community-driven vs Anthropic-built).
Best for: technical captains who want a local-first, open-source agent under their own control.
Skip unless you're technical — but for hackers, this is an excellent free Cowork alternative.
C
Agentic Frontier
Local code-execution agent
Run code on your machine via natural language.
AgenticVery High
MCPYes
Open-source. Tell it 'organize my downloads folder' or 'analyze this CSV'; it writes and executes Python on your laptop. The original 'AI runs my computer' tool — still going strong.
Free, open source · You provide your own model API key
Redundancy check — Overlaps Goose and Cowork. Open Interpreter is the most code-execution-focused of the three.
Best for: technical captains who want to run agentic code on local files without sending them to the cloud.
For the captain doing data work or local automation — useful, free, but technical.
C
Agentic Frontier
Autonomous research / task agent
Multi-agent system for research, presentations, and complex tasks.
AgenticHigh
MCPNo
Genspark Super Agent runs orchestrated AI subagents to deliver research reports, build presentations, do market scans. Strong for captains who want one tool to handle full project workflows.
Free tier · Plus $19.99/mo · Pro $44.99/mo
Redundancy check — Comparable to Manus. Genspark is more polished, Manus is more raw / flexible.
Best for: captains who want a research-and-build agent for ministry, business, or content tasks.
Free tier shows what's possible. Plus tier is reasonable if you do project work weekly.
C
Agentic Frontier
Open-source browser automation agent
Open-source agent that drives any website.
AgenticVery High
MCPPlugin
Lets any LLM (Claude, GPT, Gemini, etc.) operate a real Chrome browser. Open source, fast-moving. The technical foundation many newer agents are built on.
Free, open source · You provide your own model API key
Redundancy check — Different shape than Comet / ChatGPT Agent (BYOM library vs polished product).
Best for: technical captains who want to script their own browser-driving agents.
Skip unless you're technical and want to build agents. Otherwise use Comet / ChatGPT Agent.
No tools in this filter.