I Tested OpenAI's Dots vs. Meta's Muse. What You Need to Know.

Product Comparison (Dots by OpenAI vs. Meta Muse)

What the video is about

The reviewer compares OpenAI “Dots” and Meta “Muse” across:

• Features/capabilities
• Integrations
• Pricing
• User experience (UX)
• Overall product direction

---

Key features mentioned

Muse (Meta)

Core concept

• An always-on “resident” cloud agent with:
  • Scheduling
  • A cloud computer

Memory & identity controls

• Memory & identity are managed via editable files:
  • “Memory file” (facts, preferences, commitments)
  • Persona / “soul” file (identity/personal style)

Chat & threads

• Multiple chat threads, including:
  • A main chat
  • Additional topic-specific chats

WhatsApp connector

• Muse can respond in real time on WhatsApp
• The Muse app is mainly for viewing that interaction; it may not support responding from that view.

Meta ecosystem positioning

• Strong “Meta ecosystem” feel, with references to:
  • Instagram
  • Facebook
  • Facebook Marketplace
• Positioned as less “developer/AI tool” oriented.

Artifacts & media creation

• Can create:
  • Documents
  • Presentations
  • Interactive dashboards
• Can generate media such as:
  • Images
  • Videos
  • Podcasts (e.g., turning news into a morning podcast)

Ideas/goals feed

• Includes goal categories like:
  • sleep, health, relationships, career, interests
• Generates:
  • ideas
  • artifacts
• Feed can be shaped via likes to tune recommendations.

UX features

• “Approvals” tab
• “Schedule” tab with reminders/tasks and a “heartbeat” status (about every ~30 minutes)
  • Checks Google Docs/email
  • Sends notifications/reminders
• Search across chats
• Reviewer characterizes it as simpler than multi-bot systems.

Model

• Runs on Muse Spark 1.3
• Reviewer view:
  • Very good for professional/intellectual work
  • “Pretty comfortable” even for non-technical users

---

Dots (OpenAI / ChatGPT ecosystem)

Core concept

• Also described as an always-on resident cloud agent with:
  • Scheduling
  • A cloud computer

Integration-first design

• Designed for users already in the ChatGPT/Codex workflow
• Uses direct access to ChatGPT/Codex after granting access
• Emphasizes an ecosystem of connectors
  • Easier than manually managing many API keys
• Can run scripts and connect APIs even if a connector doesn’t exist yet.

Communication style / agent model

• Reviewer expected “multiple bots” (like Grok), but says Dots feels more like:
  • one agent / one interaction stream (similar to Muse)

Output artifacts

• Produces results in formats like:
  • Markdown
  • Dashboards
• Leans on ChatGPT memory first, then builds additional context over time.

Communication options

• Phone support (app equivalents are referenced)
• Emphasized differences vs. Muse (video focuses on those).

Slack integration

• Presented as a major strength:
  • Add the app in “three clicks”
  • Supports threaded chat
  • Supports team messaging
  • Shows a “thinking” state while responding

Calling / voice

• Claims Dots can be “called” and has low-latency voice
• Reviewer notes some prior demo issues earlier.

Team/workflow positioning

• “Direct path to Codex and Slack” is highlighted as a major selling point.

Model

• Runs on GBD-6 Astra
• Reviewer’s rough characterization:
  • More intelligent
  • Better for coding, app/software building, and automation
  • Preferred for:
    • startup scripts
    • automations
    • dashboards (coding-heavy tasks)

---

Pricing details mentioned

Muse

• Free tier: up to 100 million tokens/week
• Paid subscription available to increase token limits

Dots

• Requires a paid ChatGPT subscription to create your first “dot”
• Mentions a promotion where current dot usage may not count against weekly limits
  • Reviewer expects this may change later

---

Pros / Cons highlighted

Muse — Pros

• Best overall UX, especially for non-technical people
• Editable memory through a clear interface (“memory file” / “soul persona”)
• Strong organization UX:
  • approvals
  • schedule
  • heartbeat checks
  • chat search
• Supports WhatsApp, intended for low-friction everyday use
• Strong artifact + media + ideas capabilities (more engaging for beginners)

Muse — Cons (implied/limited)

• Not described as the top choice for heavy coding
• Positioned as less “developer-forward” than Dots/Astra

---

Dots — Pros

• Best fit if you already use ChatGPT + Codex
• Slack integration and team workflow fit
• Leverages the existing ecosystem (“local skills”)
• Better for coding/automation (Astra model comparison)
• “Calls”/voice and real-time responses mentioned as strengths

Dots — Cons

• Reviewer felt it was rushed and buggier
• Scheduling management issues:
  • told to use a “scheduled section,” but it was missing
• Organizational issues:
  • tasks/threads generated but not appearing in the expected project/organization (example mentions “Herk too”)
• Interface described as bad/abstract/rushed
• Less control over memory/profile than Muse

---

Comparisons made in the video

• Versus Grok
  • Reviewer expected multi-bot behavior, but says Dots is more like Muse:
    • a single stream/agent
• Versus Codex/ChatGPT ecosystem
  • Dots is framed as the natural extension if you already work there.
• Convergence prediction
  • Reviewer predicts Muse/Grokbot/Dots will merge functionally over time, with differences mainly in:
    • integrations/connections
    • underlying model
    • price tiers

---

Overall user experience summary

• Muse: polished, intuitive, and “easy to evaluate for free,” with strong organization and consumer-friendly communication (notably WhatsApp).
• Dots: powerful for development/automation, but in practice feels less stable, less organized, and incomplete/rushed.

---

Unique points list (all mentioned)

1. The video compares Dots vs. Muse on features, capabilities, pricing, and direction.
2. Muse reached top position in Apple’s App Store (claimed).
3. Dots felt rushed/copycat to catch up after Muse’s rise.
4. Dots demo issues/crashes were noted during Dev Day.
5. Both support naming and some agent customization.
6. Both are resident cloud agents with scheduling and cloud computers.
7. Models: Dots runs on GBD-6 Astra; Muse runs on Muse Spark 1.3.
8. Dots offers direct access integration with ChatGPT/Codex after granting access.
9. Both have their own computer/tools; both use connectors and simpler SSO.
10. Both can run scripts and connect external APIs beyond built-in connectors.
11. Both can create documents/presentations/dashboards stored in their ecosystems.
12. Dots starts with ChatGPT memory and expands context over time.
13. Muse supports editable “memory file” plus visual identity controls (“soul/persona file”).
14. Muse interface is considered better for non-technical users.
15. Both support phone communication; Dots integrates quickly with Slack.
16. Slack: Dots shows “thinking,” replies in threads, and supports team messaging.
17. Muse includes WhatsApp interactions; Muse app viewing doesn’t necessarily allow responding from the viewing thread.
18. Ecosystem alignment differs: Dots favors programming ecosystem; Muse aligns with Meta social ecosystem.
19. Muse free tier: up to 100M tokens/week; Dots needs paid ChatGPT.
20. Dots may have a promotion where dot usage doesn’t count (reviewer expects later change).
21. Model comparison: Astra is “clearly better” (at least at the time of review).
22. Astra expected to be better for coding/app building/automation; Spark may be better for general assistant tasks.
23. Muse UX includes “one thread,” approvals tab, schedule tab, and heartbeat checks (~30 minutes).
24. Muse schedule checks Google Docs/email and sends reminders.
25. Muse includes chat search and simpler chat management without many bots.
26. Muse includes an “ideas/goals” feed with personalization via likes.
27. Muse can generate images/videos/podcasts.
28. Dots advantage: “call/talk” with low latency and voice (noting earlier demo issues).
29. Dots scheduling management is confusing; scheduled section couldn’t be found.
30. Dots organizational issue: created tasks/threads didn’t post where expected in the organization/project.
31. Dots described as buggy; reviewer joked it should be called “Bugs.”
32. Reviewer would recommend Muse to non-technical parents/average users.
33. Reviewer would choose Dots if the user is already working in Codex/Claude Code and wants automation/coding.
34. Prediction: Muse/Grokbot/Dots converge into one tool experience; differences remain in integrations/model/price.
35. Reviewer suggests Astra is clearly better than Muse Spark (at least at that point).
36. Reviewer discusses tradeoffs between multi-bot organization (Grokbot-like) vs single-chain agent model.
37. Final recommendation: experiment with both; choose based on whether your workflow is ChatGPT/Codex (Dots) or consumer-friendly assistant workflow (Muse).
38. Outro includes a prompt to like/subscribe (not product-related).

---

Speaker-specific views

• Single main reviewer perspective throughout:
  • Strong emphasis on Muse UX/persona/memory/organization
  • Balanced with Dots integration + coding potential
  • Also highlights Dots bug/rushed/scheduling issues
• No separate named speakers with distinct opinions were provided.