🔍 Read the full analysis: A Closer Look At The 5X Subsidy Inside AI Subscriptions on ThorstenMeyerAI.com
Get the little things that make your day delivered free — and shop member deals
- Fast, free delivery on millions of items
- Access to Prime Big Deal Days deals on October 6–7
- Prime Video, Amazon Music and more included
TL;DR
SemiAnalysis measured usage limits across AI subscriptions and estimated that Claude plans deliver about 5.4 to 5.6 times the API-priced usage of comparable ChatGPT plans on a coding-agent workload. The report also describes recent limit changes and estimates that heavy use of premium models can make subscriptions costly for providers. The figures depend on the tested workload, model prices and plan limits.
SemiAnalysis measured how much usage major AI subscriptions provide and estimated that, for a coding-agent workload, Claude plans deliver about 5.4 to 5.6 times the API-priced usage of comparable ChatGPT plans at the same monthly price. The comparison covers specified models and plan limits, and comes as OpenAI has cut allowances on its $200 plan and introduced a $500 tier.
The report tests how each provider’s usage bar moves for different token types, then prices the measured usage at first-party API list rates. It calls that estimate “API value”: the cost of a plan’s full monthly usage limit if the same tokens were bought through the API. Its main side-by-side comparison uses a coding-agent workload dominated by cached input: the report estimates 96.6% cached input, 2.6% cache writes, 0.4% fresh input and 0.3% output.
At $20 a month, SemiAnalysis estimates $211 in API value for ChatGPT Plus using GPT-6.1 Sol, compared with $1,178 for Claude Pro using Opus 5.5. At $100, its estimates are $1,055 for ChatGPT Pro 100 and $5,725 for Claude Max 5x. At $200, they are $2,084 for ChatGPT Pro 200 and $11,726 for Claude Max 20x. Those figures produce reported ratios of about 5.4 to 5.6 times. The report says the gap also remains large when comparing raw token counts, which avoids some effects of different API prices.
SemiAnalysis says OpenAI recently cut usage allowances on its $200 plan roughly in half. New buyers get the lower limits immediately; existing subscribers keep their previous limits until October 29. The company also added a $500 plan. According to the report, that tier provides about 21% more Astra usage than the former $200 plan, but less Sol-class API value after GPT-6.1 Sol’s cached-input price cut. The report identifies a 300-token-per-second “Ultrafast” mode as the new tier’s main selling point and says it is still testing that feature.
The 5x is a subsidy, not a price
SemiAnalysis metered the meters — every major AI subscription, token type by token type, converted to API list value. On the mid-tier models both labs call the daily driver, a Claude plan returns ~5–6× the API value of the matching ChatGPT plan. Real — and the least durable number in the report.
…and the plan is fully exhausted. One pool for every model.
…and the plan is only half used — Fable is capped at 50% of the limit, leaving the rest for Opus/Sonnet. That’s where the mid-tier gap compounds.
- $200 plan halved — Sol-class value down >50% (6.1 Sol cache price cut compounds it)
- Old limits kept until 29 October; new buyers cut immediately
- New $500 tier: only +21% Astra vs the old $200 — real draw is 300 TPS Ultrafast
- Ladder flattened: Pro 100/200/500 now identical per dollar; multipliers removed from pricing page
- In OpenAI’s favour: no 5-hour window on Pro plans — easier to use the full allowance
- Flat per-dollar value across all tiers, before and after
- New premium models placed at lower relative limits (Fable capped at 50%)
- Opus allowances raised ~20% (Max) / ~50% (Pro) with the 5.5 price cut — not enough to fully offset it
- Repeatedly walked back planned cuts earlier this year under pressure from OpenAI’s generosity
- Twelve months ago, OpenAI was the generous option. Positions swap.
Gross margin per plan, assuming 92% API gross margins. The subsidy lives almost entirely in Opus and Sonnet usage — Anthropic would already be near software-like subscription margins if everyone used only Fable. Subscriptions matter even more for OpenAI, where they’re a larger share of revenue.
Three identical subscriptions; one had ~20% lower limits. The provider (unnamed) confirmed an “extremely tiny” A/B test on limit balancing. Two lessons: limits can change silently, per account, at any time — and you won’t know without instrumentation. The usage bar is a percentage, not a contract.
If you’re choosing a plan this month for agentic coding on a mid-tier model, the report settles it: a Claude plan returns ~5–6× the API value of the matching ChatGPT plan. But a plan returning 58× its fee on a model served at a steeply negative margin for heavy users is a marketing budget with a usage meter. Value moves silently, gets A/B tested per account, and twelve months ago ran the other way. Use the subsidy while it exists — it’s genuinely large. Don’t build a cost model on it. Price workloads at API rates, keep a router between you and any one vendor, and benchmark open weights on your own hardware for steady volume. A deal you can’t verify isn’t a price. It’s weather.
The Cost of Heavy Model Use
The comparison matters because subscription limits affect both what customers can use and what providers earn for their computing capacity. SemiAnalysis estimates that subscriptions account for about 10% of Anthropic revenue while consuming more than 40% of its inference compute. It estimates that this lowers blended revenue per megawatt by about $36 million. The report says subscriptions represent a larger share of OpenAI revenue, though it does not give a corresponding share in the supplied figures.
The report’s margin estimates show how strongly the economics depend on customer usage and model choice. Assuming a subscriber uses the full allowance and API business gross margins are 92%, SemiAnalysis calculates an estimated gross margin of about −369% for a fully used Opus 5.5 plan and about 1% for a fully used Fable 5.1 plan. At 20% average utilization, its estimates rise to about 6% and 80%, respectively. These are estimates under stated assumptions, not reported results from either company’s accounts.
For customers, the practical value depends on how much they use a plan, which models they select and how the provider measures limits. A large API-equivalent figure does not mean a subscriber receives cash or can use every model without constraints. OpenAI plans, the report says, lack a five-hour usage window, which may help people who need to use more of their monthly allowance in short bursts.
As an affiliate, we earn on qualifying purchases.
Recent Price and Limit Changes
API prices and subscription allowances have changed on both sides of the comparison. SemiAnalysis says Anthropic cut Fable 5.1 cache-read prices by 75% compared with Fable 5, while Opus 5.5 prices fell 20% for input and output and 60% for cache reads compared with Opus 5. The report says Fable 5.1 launched without higher token limits. Opus allowances rose by about 20% on Max and 50% on Pro, but the report says those increases did not fully offset the lower API prices.
It describes a similar effect for GPT-6.1 Sol: OpenAI cut the model’s cached-input price without raising its limits, reducing the plan’s estimated API value. SemiAnalysis puts the decline on the $200 plan at about 30%. In general, a lower API price reduces an API-equivalent estimate when the subscription’s token allowance stays fixed; the subscriber’s plan fee need not change for that calculation to fall.
The report also compares frontier models separately. It says GPT-6 Astra’s allowance on a $200 plan is exhausted after about $2,897 worth of API usage, while Claude Fable 5.1 uses half of a Claude plan’s limit after about $2,485 worth. The remaining half of the Claude limit is available for Opus or Sonnet, according to the report. This model-specific comparison is distinct from the Opus-based figures behind the headline ratio.
Limits Behind the Estimates
The ratios are estimates for a particular workload, model mix, metering method and set of API prices. The tested coding-agent workload is unusually heavy in cached input, so customers with different tasks or token patterns may see different practical value. The figures do not establish what a typical subscriber uses or how many customers reach their plan limits.
SemiAnalysis says the value gap persists in raw token counts, but the supplied material does not give the underlying counts or full testing methodology. The report’s margin calculations also depend on assumptions about utilization and API gross margins. The supplied account does not include company responses confirming those estimates, and it does not say how long current allowances or prices will remain in place.
Plan Changes and Further Testing
The next dated change in the report is October 29, when existing subscribers to OpenAI’s $200 plan are due to lose their grandfathered limits, according to SemiAnalysis. New purchases already receive the reduced allowance. The report also says it is still testing the new $500 plan’s Ultrafast mode, so its performance and practical value remain unsettled.
Further comparisons will depend on future model prices, subscription limits and usage rules. If providers reduce API prices without raising subscription allowances, API-equivalent values may fall even while monthly fees stay the same. Readers should treat the reported ratios as a snapshot of the tested plans and workload, rather than a guarantee of value for every subscriber.
Key Questions
What does the reported 5.6× figure measure?
It compares estimated API list-price values for the full usage limits of a $20 ChatGPT Plus plan using GPT-6.1 Sol and a $20 Claude Pro plan using Opus 5.5. SemiAnalysis estimates $211 and $1,178, respectively, for its coding-agent workload.
Does every Claude subscriber get 5.6 times more usage?
No. The estimate depends on the selected models, token types, workload, plan limits and API prices. The report’s workload is dominated by cached input, and another customer’s usage may produce a different comparison.
When do existing $200 ChatGPT Pro limits change?
SemiAnalysis says existing subscribers keep their previous limits until October 29. New purchases receive the reduced limits immediately, according to the report.
Why can an API price cut lower a subscription’s estimated value?
The report prices measured subscription usage at API list rates. If the API price per token falls while the subscription allowance stays fixed, the same allowance has a lower API-equivalent dollar value.
Source: ThorstenMeyerAI.com
Evergreen bestsellers Picks
bestsellers
As an affiliate, we earn on qualifying purchases.
