ChatGPT Pro 500 vs 200 vs 100: pricing, usage limits, Ultrafast and GPT-6.1 Sol API prices
On September 29, 2026, OpenAI split ChatGPT Pro into three monthly tiers — Pro 100, Pro 200 and the new Pro 500 (OpenAI Help Center: About ChatGPT Pro tiers). Pro 500 is the headline change: the highest usage allowance OpenAI sells on a consumer plan, 25 times the Plus allowance, and the only Pro tier with Ultrafast (OpenAI: DevDay 2026 Recap (Sep 29, 2026)). Below: the three tiers side by side, what Ultrafast actually costs, and where GPT-6.1 Sol fits in, because Sol’s API price is the other half of the Pro-500 math.
ChatGPT Pro 500 vs 200 vs 100 in one table
| Question | What OpenAI says | Official source |
|---|---|---|
| Pro 100 price | $100/month. The entry Pro tier; Ultrafast is not included. | OpenAI Help Center: About ChatGPT Pro tiers |
| Pro 200 price | $200/month. Reopened for new subscriptions on Sep 29, 2026 with a lower included allowance than before; Ultrafast is not included. | OpenAI Help Center: About ChatGPT Pro tiers |
| Pro 500 price | $500/month. The highest usage allowance of the three tiers — 25 times the ChatGPT Plus allowance — and the only Pro plan that includes Ultrafast. | OpenAI Help Center: About ChatGPT Pro tiersOpenAI: DevDay 2026 Recap (Sep 29, 2026) |
| How the included allowances compare | OpenAI publishes multiples, not message counts: Pro 200 includes more usage than Pro 100, and Pro 500 has the highest allowance (25× Plus). Exact per-model counts are not published and vary by model. | OpenAI Help Center: About ChatGPT Pro tiersOpenAI: DevDay 2026 Recap (Sep 29, 2026) |
| How Ultrafast is billed on Pro 500 | Ultrafast draws on the plan’s included usage first, then the credit balance. Buying credits on Pro 100 or Pro 200 does not unlock Ultrafast at launch. | OpenAI Help Center: About ChatGPT Pro tiers |
| Going past your included usage | Included usage is used first, then credits. Credits cover eligible usage in Codex, ChatGPT Work and ChatGPT for Word, Excel and PowerPoint — not every model or feature. | OpenAI Help Center: About ChatGPT Pro tiers |
| Existing Pro 200 subscribers | Eligible subscribers keep their previous included allowance through Oct 29, 2026, then move to the lower new-subscriber allowance at the same $200 price. Grandfathering does not add Ultrafast or any Pro 500 feature. | OpenAI Help Center: About ChatGPT Pro tiers |
| When a model hits its limit | ChatGPT displays the reset time when available. Support cannot reset ChatGPT or Codex limits, and no setting can bypass or raise a limit. | OpenAI Help Center: About ChatGPT Pro tiers |
Pro 500 vs 200 vs 100: how to read the tiers
The practical difference between the tiers is not features — all three include Pro models, Codex, deep research, image creation, memory and file uploads (OpenAI Help Center: About ChatGPT Pro tiers). The difference is how much agentic work you can run before waiting. Pro 200 roughly doubles Pro 100 on price and includes more usage; Pro 500 is the step-change tier, aimed at people who run long Codex or ChatGPT Work sessions daily and would otherwise hit walls mid-week.
Two details matter more than the sticker price. First, Pro plans currently have no five-hour limit in Codex — the five-hour windows that shape Plus and Business usage don’t apply to Pro, so the allowance is the constraint, not the clock (OpenAI Learn: Codex pricing (usage, credits and speed modes)). Second, the Pro 200 grandfathering cliff: after Oct 29, 2026, new and re-subscribed Pro 200 accounts carry a noticeably lower allowance than legacy ones (OpenAI Help Center: About ChatGPT Pro tiers). If you are comparing Pro 500 vs 200 for a purchase made after that date, compare against the new Pro 200 allowance, not the legacy one.
Astra Ultrafast pricing: 8× speed, 6× price
Ultrafast is OpenAI’s premium speed tier: up to 8× faster token generation (300 tokens per second) in Codex and ChatGPT Work, and up to 6× in the API (OpenAI: DevDay 2026 Recap (Sep 29, 2026)). On subscriptions it is billed at 8× the standard subscription usage rate; purchased credits and Enterprise pay-as-you-go spend at 6× the standard credit rate (OpenAI Learn: Codex pricing (usage, credits and speed modes)). Among Pro plans it exists only on Pro 500 — and there it spends included usage first, credits after (OpenAI Help Center: About ChatGPT Pro tiers).
| API price per 1M tokens | Input | Cached input | Cache writes | Output |
|---|---|---|---|---|
| GPT-6 Astra, Standard (short context) | $10.00 | $1.00 | $12.50 | $50.00 |
| GPT-6 Astra, Standard (long context) | $20.00 | $2.00 | $25.00 | $75.00 |
| GPT-6 Astra, Ultrafast (short context) | $60.00 | $6.00 | $75.00 | $300.00 |
| GPT-6 Astra, Ultrafast (long context) | $120.00 | $12.00 | $150.00 | $450.00 |
Prices from the OpenAI API pricing table (OpenAI API Docs: API pricing (per 1M tokens)). In the API you opt in per request with service_tier: "ultrafast"; the mode is broadly available for GPT-6 Astra, with GPT-6.1 Sol Ultrafast announced as coming soon (OpenAI API Docs: Ultrafast mode guide OpenAI: DevDay 2026 Recap (Sep 29, 2026)). The multiplier is on tokens, not on speed — a reply that takes a third of the time still costs six times the standard token price.
GPT-6.1 Sol pricing: near-Astra quality at one-fifth of the API price
GPT-6.1 Sol is the budget side of the Pro-500 story. OpenAI’s announcement prices it at one-fifth of Astra’s standard input and output token prices, with cached input at $0.10 per million tokens — 95% below standard input pricing (OpenAI: Introducing GPT-6.1 Sol (Sep 29, 2026)). The API pricing table confirms the exact figures:
| API price per 1M tokens | Input | Cached input | Cache writes | Output |
|---|---|---|---|---|
| GPT-6.1 Sol, Standard (short context) | $2.00 | $0.10 | $2.50 | $10.00 |
| GPT-6.1 Sol, Standard (long context) | $4.00 | $0.20 | $5.00 | $15.00 |
| GPT-6 Astra, Standard (short context) — for comparison | $10.00 | $1.00 | $12.50 | $50.00 |
OpenAI API Docs: API pricing (per 1M tokens) OpenAI: Introducing GPT-6.1 Sol (Sep 29, 2026). In ChatGPT apps, GPT-6.1 Sol is available in ChatGPT Work and Codex — not in regular Chat — on paid plans from Plus up (OpenAI: DevDay 2026 Recap (Sep 29, 2026)). Note the “GPT-6.1 Sol vs Astra” question cuts both ways in subscription usage: Astra answers at the frontier, Sol answers at roughly a fifth of the cost, and on Plus and Business the five-hour message estimates reflect that — Sol gets roughly 15–160 local messages per five hours where Astra gets 5–45 (OpenAI Learn: Codex pricing (usage, credits and speed modes)).
Track the reset clocks Pro doesn’t remove
Pro removes the five-hour wall but not the allowance itself: per-model limits still reset on their own schedule, ChatGPT shows the reset time when one is available, and Support cannot reset limits for you (OpenAI Help Center: About ChatGPT Pro tiers). QuotaClock turns those reset times into a countdown you can leave open:
- Read the reset time in ChatGPT or Codex (Settings → Usage in ChatGPT, Usage & Billing in Codex — or the usage dashboard).
- Register it in the countdown tool: set Window started at to now, type the minutes until reset as the window length, and name it after the model.
- Confirm at zero. A zero only means your arithmetic ran out — check ChatGPT before assuming the allowance is back.
Running Codex alongside Claude Code? The Claude Code limit reset guide covers Anthropic’s reset clocks, and the ChatGPT dots limits page covers what dots add to a Pro subscription. If Codex usage is the thing you keep hitting, how Codex Cloud spends your allowance and what Codex Security Cloud costs are the two pages that matter.
Frequently asked questions
Is ChatGPT Pro 500 worth it over Pro 200?
It depends on whether you actually hit Pro 200’s walls. The objective differences: Pro 500 has the highest allowance (25× Plus), it is the only Pro tier with Ultrafast, and it has no five-hour limit in Codex — but Pro tiers in general carry no five-hour limit, so that last point applies to Pro 100 and 200 as well. OpenAI publishes no per-model message counts, so the honest test is a week on your current tier.
Does Pro 200 still exist for new subscribers?
Yes. Pro 200 reopened for new subscriptions on Sep 29, 2026 at $200/month, but new subscriptions that aren’t eligible for grandfathering include a lower usage allowance than the old Pro 200. Eligible existing subscribers keep the previous allowance through Oct 29, 2026.
Can I buy Ultrafast on Pro 100 or Pro 200?
No. At launch, buying credits on Pro 100 or Pro 200 does not unlock Ultrafast. Among Pro plans it is available only on Pro 500, where it spends included usage first and then credits.
What are the GPT-6.1 Sol API prices?
Per 1M tokens, standard short-context pricing is $2.00 input, $0.10 cached input, $2.50 cache writes and $10.00 output — one-fifth of GPT-6 Astra’s standard input and output prices. Long-context pricing is $4.00 / $0.20 / $5.00 / $15.00.
What happens when I hit my Pro usage limit?
The model shows as unavailable until its allowance resets, and ChatGPT displays the reset time when one is available. You can switch to another available model, wait for the reset, or — for eligible features in Codex, ChatGPT Work and the Office add-ins — buy credits to continue. Support cannot reset the limit manually.
Is there an annual ChatGPT Pro price?
The Pro tiers OpenAI documents are monthly — $100, $200 and $500 per month. The pages cited above give no annual Pro pricing, so this page doesn’t list one.
Sources
All official OpenAI pages, last read 2026-09-30:
- OpenAI: DevDay 2026 Recap (Sep 29, 2026)
- OpenAI Help Center: About ChatGPT Pro tiers
- OpenAI API Docs: API pricing (per 1M tokens)
- OpenAI: Introducing GPT-6.1 Sol (Sep 29, 2026)
- OpenAI Learn: Codex pricing (usage, credits and speed modes)
- OpenAI API Docs: Ultrafast mode guide
QuotaClock is not affiliated with, endorsed by, or connected to OpenAI. “ChatGPT”, “Codex”, “GPT-6” and “Ultrafast” are OpenAI product names, used here only to describe them.
QuotaClock is an independent, non-official tool. It is not affiliated with, endorsed by, or connected to Anthropic, OpenAI, or any other AI provider. It does not sign in to your accounts, does not read API keys, tokens, or usage data, cannot reset anything for you, and cannot promise that a provider's window rules stay the same. The countdown is built only from the times and lengths you type in. Window lengths and reset rules are set by each provider and change over time — always check the provider's own current documentation.