CAVEMAN NEED CODEX NOW
Calculating your bridge
Using the 14-point/day fallback until you edit it.
* 14/day fallback = 100 ÷ 7 ≈ 14.29Traffic: loading…
Huh?
days left × 14 points/day − 7 points = your gap
gap × current plan capacity = normalized gap
same observed token mix × normalized gap × official rates
official API × ½ = OpenRouter Aug 17 discount
show cheapest reviewed plan with normalized capacity ≥ gap
covering higher-plan upgrade wins an equal-cost bridge tie
The $100 plan is the default because it matches the observed cohort. Changing Current plan scales the same percentage gap through reviewed plan capacity. A covering higher-capacity plan wins when its incremental upgrade cost is no more than the otherwise cheapest bridge; this is planning guidance only, not a purchase or guaranteed limit. Options sort cheapest first. The translucent red X crosses the losing method, not its number.
Still here? You not caveman. Fine.
Numbers actually used
- Sample
- 397,415,420 emitted tokens while remaining quota moved 55% → 7% across 48 points.
- Token mix
- 10,579,239 uncached input · 385,656,320 cached input · 1,179,861 output.
- Credits
- Sol: 125 input · 12.5 cached · 750 output credits per 1M tokens.
- API
- Sol: $5 input · $0.50 cached · $30 output per 1M short-context tokens.
- Observed
- to ; compactions occurred inside the sweep.
- Rates
- .
- OpenRouter
- OpenRouter's says GPT-5.6 Sol is half off across batch, flex, and priority tiers; no expiry is stated.
Subscription rule
Current-plan choices with comparable capacity are Plus $20 (20 normalized points), Pro 5x $100 (100, the default observed cohort), and Pro 20x $200 (400). Go $8 is reviewed but unavailable as a current-plan choice because comparable Sol capacity is not published. Cavedex converts the entered percentage gap through the chosen plan's normalized capacity, then shows only the cheapest tier whose capacity covers that normalized gap. When a higher tier adds enough capacity to cover the gap and its incremental monthly cost is no more than the otherwise cheapest bridge, the upgrade wins the tie; for Pro 5x to Pro 20x that means $200 total, +$100 incremental, and +300 normalized points. No purchase or account change occurs. These ratios follow the official Plus/5x/20x labels; task complexity and limits vary, so capacity is planning guidance, not a guarantee.
OpenRouter source boundary
Discounted Sol query · OpenRouter status. The status says flex can be as low as $1.25 input and $7.50 output. * Flex is a separately selected pricing tier and the status does not state cached-input treatment, so the compact comparison conservatively uses the attributable half-off standard-token total. OpenRouter is a third-party source, the promotion may expire, and eligibility must be rechecked.
Receipts
Usage observation · Official credits, plans, and limits · Official API rates.
The trace may miss other account use. Compactions occurred inside the 48-point sweep, and quota conversion remains magnitude-only rather than universal. Rates, plan eligibility, and third-party promotion status need rechecking.