Why Prices Vary by Model and Quality Tier Instead of One Flat Rate

October 11, 2026 · Soulris Team · 2 min read · chat, images, video, wallet

There's no single flat price

Ask SoulrisAI a quick question and generate a high-end video in the same session, and you'll pay very different amounts for each. That's intentional. A flat, one-size-fits-all price would mean light, cheap requests subsidizing expensive ones — or, if priced to cover the expensive end, overcharging every simple request to do it.

What actually drives cost

Three things determine what a given request costs:

  1. Which model answers it. Different AI models genuinely cost different amounts to run. A higher chat tier — Intelligent, Expert, or Super — is more capable, and that extra capability costs more to provide than Basic.
  2. How much work the request involves. A short message costs less than a long one; a higher image quality tier or a longer, higher-resolution video costs more than a shorter, simpler one. See video quality tiers explained for how that plays out specifically for video.
  3. What actually happened, not what might have. You're charged for the real cost of what was generated, not a flat estimate — see holds vs. settlement explained for exactly how that's calculated and charged.

Why Auto routing matters here

For chat, you don't usually need to think about any of this directly. Auto routing tries to match each message to the lowest tier that can genuinely handle it well, rather than defaulting everything to the most expensive tier. That's the main lever that keeps your everyday chat costs down without you having to manage it yourself.

Why this is fairer than one flat rate

Under a flat rate, someone who mostly asks simple questions pays the same as someone running demanding, expensive requests constantly. Tying price to actual cost means you pay roughly what you use, not a number averaged across everyone else's usage.

Can I see the price before I commit?

For image and video generation, yes — the Studio shows you a cost estimate before you generate, so you're never guessing. See estimating video cost before you generate. For chat, the cost of an individual message is small enough, and routing handles tier selection well enough, that most users don't need to check a price before every message — though your wallet balance is always visible if you want to keep an eye on it.

This is also why SoulrisAI isn't a flat subscription

See pay-per-use vs. subscription for how this cost-reflects-usage approach compares to paying one fixed monthly amount regardless of what you actually use.