01 / Pricing that gives you room
Every plan includes the full model catalog. Choose monthly capacity from $20—or try the API with $2 prepaid credits. No overage billing.
Explore pay as you goOne OpenAI-compatible key · Secure Stripe checkout · Requests pause at your limit
02 / Monthly plans
Choose by usage. Every plan pauses at its limit instead of charging overage.
A focused monthly lane for individual builders.
Best first monthly plan
1x monthly capacity
Generous usage limits, reset monthly
Secure Stripe checkout · Cancel anytime · No overage billing
For daily agent work, deeper refactors, and longer sessions.
Best for daily agent work
5x monthly capacity
Generous usage limits, reset monthly
Secure Stripe checkout · Cancel anytime · No overage billing
The high-capacity lane for launches and recurring agent runs.
Best for launches and heavy runs
20x monthly capacity
Generous usage limits, reset monthly
Secure Stripe checkout · Cancel anytime · No overage billing
Maximum public capacity for teams and always-on agent workloads.
Best for teams and always-on agents
50x monthly capacity
Generous usage limits, reset monthly
Secure Stripe checkout · Cancel anytime · No overage billing
Capacity is supplier-weighted: premium models consume allowance faster than lighter models. All monthly allowances reset each billing cycle and do not roll over.
Hard caps pause requests instead of charging extra.
Your API key works immediately after checkout.
Prepaid balance becomes a Stripe discount on upgrades.
Track tokens, spend, and allowance in real time.
03 / Find your fit
You are never choosing model access—all four plans include the complete public catalog. You are only choosing room to run.
Jump to plansA few focused sessions each week
Start here if you are learning your usage.
Daily coding and deeper refactors
The best default for serious solo builders.
Launches and recurring multi-agent runs
Choose it when work arrives in sustained bursts.
Team-wide or always-on workloads
Maximum public capacity with the same hard cap.
04 / No subscription needed
Prefer flexibility? Prepaid credits pay for successful inference only — one-time payment, no auto-refill, credits last 12 months, and usage simply pauses at zero.
Prepaid balance becomes a Stripe discount when you move to a monthly plan.
One-time payment
$2USD
Secure Stripe checkout · No automatic refill
05 / The receipt
GPT-6 PAYG costs $5 input, $0.50 cached input, and $25 output per 1M tokens versus $10 / $1 / $50 for OpenAI gpt-6-astra Standard short-context pricing, checked September 16, 2026. The 50% comparison is not against Batch, Flex, Fast mode, long-context rates, or monthly packages; cache writes, tools, and regional charges are excluded. Other models have different savings. Compare exact input, cached-input, and output rates in the live pricing API before choosing a model.
USD per 1M uncached input / output tokens. Scroll horizontally to see all rates on smaller screens.
| Model | Xpersona input / output | Official standard | You save |
|---|---|---|---|
| GPT-6 | $5 / $25 | $10 / $50 | 50% / 50% |
| GPT-5.6 / Sol | $1.50 / $12 | $5 / $30 | 70% / 60% |
| GPT-5.6 Terra | $1.50 / $2 | $2.50 / $15 | 40% / 87% |
| GPT-5.5 | $1.50 / $12 | $5 / $30 | 70% / 60% |
| GPT-5.4 | $0.75 / $6 | $2.50 / $15 | 70% / 60% |
| GPT-5.4 Mini | $0.375 / $4 | $0.75 / $4.50 | 50% / 11% |
| Claude Fable 5 | $3 / $18.50 | $10 / $50 | 70% / 63% |
| Claude Opus 4.8 | $1.50 / $9.25 | $5 / $25 | 70% / 63% |
| Claude Sonnet 4.6 | $0.90 / $5.55 | $3 / $15 | 70% / 63% |
| Claude Haiku 4.5 | $0.60 / $3.70 | $1 / $5 | 40% / 26% |
USD per 1M uncached input / output tokens. GPT-6 uses OpenAI Standard short-context pricing checked September 16, 2026 (source). Other rows retain the July 28, 2026 snapshot, not a current-price quote. Generic GPT-5.6 is compared with GPT-5.6 Sol. Batch, flex, priority, cache, long-context, tool, and regional charges can differ.
06 / The full catalog, together
Every monthly package unlocks the full catalog. Included capacity is supplier-weighted, so premium models consume the allowance faster than lighter models.
GPT-6 / GPT-5.6 / GPT-5.6 Sol / GPT-5.6 Terra / GPT-5.5 / GPT-5.4 / GPT-5.4 Mini
Frontier reasoning, agentic coding, balanced everyday work, and economical high-volume runs.
Claude Fable 5 / Opus 4.8 / Sonnet 4.6 / Haiku 4.5
Deep analysis, premium coding, balanced intelligence, and fast responses through genuine upstream routes.
Gemini 3.5 Flash / Frieren 1 / A_BETTER_YOU
Fast multimodal reasoning plus Xpersona's own coding and extra-high-reasoning lanes.
07 / From plan to first request
Flat monthly package for predictable spend, or $2 prepaid to start light.
OpenCode, Chat, or your own stack via the OpenAI-compatible endpoint — one key, every model.
Hit the pause line and decide: bigger package or prepaid credits. Your tools never reconnect.
08 / The details
Model access, monthly limits, and what happens to your credits.
Yes. Builder, Pro, Studio, and Scale unlock the same GPT, Claude, Gemini, and Xpersona lineup. Premium models consume the included supplier-weighted capacity faster than lighter models.
Builder, Pro, Studio, and Scale scale the monthly capacity 1x, 5x, 20x, and 50x on the base-cost route. Premium models draw from the same supplier-weighted compute allowance faster. Allowances reset monthly and do not roll over.
Package requests pause instead of creating surprise overage charges. You can move to a larger package or add prepaid credits.
They roll over automatically. When you check out for any monthly plan, your available credit balance is applied as a discount shown in Stripe before you pay.
Choose $2, $5, $10, $25, $50, or $100. It is a one-time payment with no subscription or automatic refill. Credits pay for successful inference, last 12 months, and usage pauses at zero.
09 / Your next build starts here
Choose Pro for daily agent work, or prove the workflow with $2 first.
