OpenAI opens the Pro plan to new users, but halves API credits and moves toward payment for actual usage
From 29. 9. 2026, OpenAI is reopening the Pro plan at $200/month to new users and removing the 5-hour usage limit, but at the same time halving API credits per dollar. The company is thus moving from subsidized subscriptions toward payment based on actual usage. This follows the price reduction for the GPT-6 Sol and Luna models on 22. 9.
From 29 September 2026, OpenAI is reopening the Pro plan at $200 per month to new users and removing the existing 5-hour usage limit — subscribers can now spread their weekly allotment however they choose. At the same time, however, the company is halving the number of API credits a user receives per dollar. According to employee Thibault Sottiaux, subscribers should nevertheless be able to get more work done than a month ago thanks to the cheaper GPT-6 Sol and GPT-6 Luna models. According to Sottiaux, this is a step away from heavily subsidized subscriptions toward payment based on actual usage (pay-per-use) — according to the source, Microsoft is taking a similar approach with part of its Copilot offering.
This development follows the release of the GPT-6 Sol and GPT-6 Luna models on 22 September 2026, which OpenAI priced 50 % lower than the previous 5.6 series. GPT-6 Sol now costs $2 per million input tokens and $10 per million output tokens, while GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens. The models are available in ChatGPT Work, Codex and the API; Luna is also available in the desktop application and to users on the Free and Go plans. According to OpenAI, GPT-6 Sol has roughly half the error rate of its predecessor on an internal factuality test, with reliability comparable to the Astra model.
OpenAI attributes the price reduction to pressure from open-source models and smaller competing offerings on frontier model prices. According to the sources, OpenAI compares GPT-6 Sol in its communications with the Claude Fable 5.1 model from Anthropic, which costs $10 per million input tokens and $50 per million output tokens — this comparison is a claim by the manufacturer, however, rather than an independent measurement. The models were released 90 minutes after the release of Claude Opus 5.5 from Anthropic, which the sources describe as a reflection of intense competition between the two companies. Analysts quoted in the sources point out that the price cuts are also linked to preparations by OpenAI for an IPO and that the price war between major providers will probably continue.
Why it matters
For companies and individuals using OpenAI products, the balance between subscriptions and usage-based payment is changing: the lower price of the new models partially offsets the reduction in API credits per dollar, but the actual value depends on the specific usage pattern. Reopening the Pro plan expands access to the higher-tier plan for new users, while companies paying through the API should recalculate whether the same usage volume remains cost-effective at the new credit conversion rate.
What was added since the original report
Verified updates
-
API credits per dollar halved; Pro plan at $200/month reopened to new users; 5-hour window for the weekly allotment removed; Strategic shift from subsidized subscription models to pay-per-use pricing; Microsoft is making similar changes with Copilot
- API credits per dollar halved
- Pro plan at $200/month reopened to new users
- 5-hour window for the weekly allotment removed
- Strategic shift from subsidized subscription models to pay-per-use pricing
- Microsoft is making similar changes with Copilot
-
Specific prices: GPT-6 Sol $2 per 1M input tokens/$10 per 1M output tokens; Specific prices: GPT-6 Luna $0.10 per 1M input tokens/$0.50 per 1M output tokens; Direct comparison with Claude Fable 5.1 as a lower-priced competitor; Explicit reason for the price reduction: pressure from open-source models and smaller models; Strategic context: the price reduction is part of the strategy ahead of the IPO
- Specific prices: GPT-6 Sol $2 per 1M input tokens/$10 per 1M output tokens
- Specific prices: GPT-6 Luna $0.10 per 1M input tokens/$0.50 per 1M output tokens
- Direct comparison with Claude Fable 5.1 as a lower-priced competitor
- Explicit reason for the price reduction: pressure from open-source models and smaller models
- Strategic context: the price reduction is part of the strategy ahead of the IPO
Two audiences, two different impacts
What this means
For individuals
Users considering the OpenAI Pro plan at $200/month now also benefit from the removal of the 5-hour limit, but those paying through the API receive fewer credits for the same dollar than before — the new lower prices of the GPT-6 Sol and Luna models only partially offset this.
For a business
Companies using the OpenAI API should recalculate their AI operating costs, because the number of credits per dollar is changing and the price of the new GPT-6 Sol and Luna models has also fallen by 50 % compared with the 5.6 series — the net impact on the budget depends on the specific usage volume and types of models used.
StrategyCheck the original
Event sources
confirmed by 3 independent sources · 3 publishers, 3 independent. We count feeds from the same owner only once.