- OpenAI Dots plans: Dots access starts at the Pro tier ($100 or $200/month); Plus users are excluded at launch
- Usage change: Pro 200 reopened with a lower usage allowance at the same $200 price
- Grace period: Existing Pro 200 subscribers keep old allowance through October 29, 2026
- GPT 6.1 Soul: Now live in working codecs across Plus and both Pro tiers, not yet in ordinary chat
- Pro 500: New tier advertised at 25x Plus usage with Ultrafast included
OpenAI Dots Plans Overview
OpenAI Dots plans determine whether you can use the always-on Dots agent, how much usage you get, and which extras like Ultrafast are included. Dots is an always-on agent powered by Aster with its own cloud computer, remembering context across the ecosystem. You can interact with it through ChatGPT, Slack, Microsoft Teams, and even phone calls and text messages (iMessage/SMS in limited rollout).
Video Highlights:
- Pro 200 reopened with reduced included usage at the same price
- Dots access starts at Pro, including $100 and $200 plans
- GPT 6.1 Soul scores 75.2 on Deep Suite at high reasoning
- First-month exemption for Dots usage, with later plan terms pending
The plan structure now spans four tiers, and the differences matter more than ever after the DevDay announcements.
| Plan | Price | Dots Access | Key Usage Detail |
|---|---|---|---|
| Plus | $20/mo | No (not yet) | Est. 15-160 local messages per 5 hours with GPT 6.1 Soul |
| Pro 100 | $100/mo | Yes | No 5-hour cap per current documentation |
| Pro 200 | $200/mo | Yes | More usage than Pro 100; allowance reduced for new/renewing subs |
| Pro 500 | $500/mo | Yes | 25x Plus usage, Ultrafast included |
The published terms explicitly confirm that Pro 200 sees a usage reduction. Eligible existing subscribers keep their old allowance through October 29, 2026, provided their $200 subscription stays active. After that, the allowance drops while the bill stays the same.
Plan-by-Plan Comparison
Each tier targets a different kind of user, and the value calculus shifted significantly after DevDay.
Plus ($20/mo)
- No Dots access at launch
- GPT 6.1 Soul available in working codecs
- Est. 15-160 local messages per 5 hours
- Local and cloudwork share the same allowance
Pro 100 ($100/mo)
- Dots access included
- No published 5-hour cap
- No fixed weekly task count documented
- No Ultrafast
Pro 200 ($200/mo)
- Dots access included
- More usage than Pro 100
- Confirmed allowance reduction
- Grace period through Oct 29, 2026
Pro 500 ($500/mo)
- 25x Plus usage
- Ultrafast included (8x faster token generation)
- Aimed at the heaviest 1% of users
- Usage impact: 8x the standard burn rate
OpenAI's explanation for reduced allowances is that models are becoming more efficient. GPT 6.1 Soul's purchase credit rates are half of GPT 6.1 Soul-free GPT 6 Soul's fresh input and output rates, and its cached input rate is a quarter. Whether that translates into more finished work for subscribers remains to be demonstrated in real-world use.
GPT 6.1 Soul Benchmark Data
GPT 6.1 Soul is the counterweight to the usage cuts. It has launched in ChatGPT working codecs, including Plus and both Pro tiers, though it is not in ordinary chat mode yet.
| Model | Deep Suite Score | Reasoning Level |
|---|---|---|
| GPT 6.1 Soul | 75.2 | High |
| GPT 6 Soul | 69.4 | High |
| GPT 6 Soul | 68.8 | Max |
Notably, GPT 6.1 Soul's max-effort score on that coding test was lower than its high-reasoning score. More thinking does not automatically buy a better result, so starting at medium or high thinking is the sensible default.
For subscription users, GPT 6.1 Soul offers two possible benefits: cheaper work and fewer retry attempts. If it finishes a job in one pass instead of several failed attempts, your allowance stretches further. These are OpenAI's published tests, so verify against your own workload.
How to Choose Your Dots Plan
Follow this decision flow before committing or upgrading.
Audit Your Current Usage
Check how much of your existing allowance you burn in a typical week. If you rarely hit limits, the Pro 200 reduction may not hurt you at all.
Decide If You Need Dots
Dots requires Pro 100 or higher. If you only need chat and coding, Plus with GPT 6.1 Soul in working codecs may suffice.
Test GPT 6.1 Soul First
Start at standard speed with medium or high thinking before assuming you need a higher tier or max effort.
Evaluate Ultrafast Value
Ultrafast gives 8x faster token generation, but that is generation speed, not 8x faster finished projects, and it burns 8x the standard rate.
Lock In Before October 29
If you are an eligible existing Pro 200 subscriber, keep your subscription active to preserve the old allowance through the grace period.
Conversations with your Dots do not count toward ChatGPT usage limits, but worker Codex tasks your Dot calls do count as usual. There is a first-month exemption for Dots usage, with later plan terms still to come.
Dots Security and Permissions
Dots ships with local computer access disabled by default, which is the right starting posture. Background proactive research is restricted to read-only connected app tools, and actions go through permissions, custom rules, and review controls.
| Control | Status | Recommendation |
|---|---|---|
| Local computer access | Optional, disabled by default | Keep off until needed |
| Background research | Read-only tools only | Review connected apps |
| Actions | Permissions + review controls | Start with narrow access |
| Text/SMS | Limited rollout | Wait for wider availability |
Take a lesson from users who lost everything with runaway agents: start with narrow access and review what your Dot produces. Controls are not bulletproof.
Dots Setup Checklist:
- Verify Pro 100 or Pro 200 subscription is active
- Keep local computer access disabled initially
- Configure custom permission rules before first task
- Review first-week Dot outputs for accuracy
- Monitor usage after the first-month exemption ends
FAQ
Q: Do OpenAI Dots plans include Plus users?
No. Dots access starts at the Pro tier, including the $100 and $200 monthly plans. Plus users do not have access at launch, though wider availability is possible as OpenAI competes with rival always-on agents.
Q: How much usage did Pro 200 lose?
OpenAI confirms a reduction in the published terms but has not published a verified percentage. Existing eligible subscribers keep their old allowance through October 29, 2026, as long as their subscription stays active.
Q: Does talking to my Dot use my ChatGPT allowance?
Conversations with your Dot do not count toward ChatGPT usage limits, but worker Codex tasks the Dot initiates count as usual. A first-month exemption for Dots usage applies, with later terms pending.
Q: Is Pro 500 worth it for OpenAI Dots plans?
Pro 500 offers 25x Plus usage and includes Ultrafast. Note that Ultrafast is 8x faster token generation, not 8x faster completed projects, and it consumes 8x the standard rate. Only heavy daily users will likely see value.