Each model has its own cost curve.
Flagship reasoning models, efficient chat models, image generation models, and video models can use different units and provider economics.
Pricing guide
AI API pricingwithoutprovider sprawl.
Understand how AllToken pricing is shaped by model choice, provider route, and usage unit, while keeping billing and cost visibility in one account.
Pricing basisModel + route + usage
Billing layerUnified account
ScopeText, vision, image, video
How pricing works
Pricing follows the route a request actually takes.
A simple text request, a multimodal request, and an async video task do not share the same cost structure. The practical way to evaluate spend is by model, route, and unit.
Provider paths can change effective cost.
A model may be available through more than one route. AllToken keeps route selection and billing in one integration surface.
Final cost follows actual usage.
Text is commonly token-based. Image and video workflows may use generated outputs, async task state, duration, or resolution.
Categories
Different modalities need different pricing mental models.
The public guide explains cost structure without pretending every live model price belongs on one static page.
TextTokens
Token-based model calls
Chat, reasoning, embeddings, and text generation are usually understood through input and output token usage.
Vision & imageRequest + output
Multimodal input and generated images
Vision requests may include image input. Image generation is usually evaluated by generated output and task settings.
VideoTask + duration
Async generation workflows
Video pricing depends on model, mode, duration, aspect ratio, resolution, and final task settlement rules.
Cost visibility
Compare the total operating model, not only a headline number.
Pricing varies by model, route, and usage. The useful comparison is how many systems, invoices, and route decisions your team has to manage.
FactorAllTokenDirect providersCredit tools
Setup overheadOne API key and one base URLSeparate accounts and onboardingTool-specific credit setup
Billing clarityUnified billing layerProvider-by-provider invoicesCredit packs can obscure unit cost
Routing controlRoute and model choice in one surfaceManual provider switchingUsually tied to one product workflow
Fallback ownershipPlan fallback paths around available routesTeams own failover logicLimited routing transparency
Cost visibilityCompare model, route, and usage before scalingRequires normalizing different provider formatsOften hard to map credits to API usage
Video pricing note
Seedance 2.0 belongs in the video cost model.
Seedance access through AllToken should be evaluated on availability, API usability, and competitive pricing visibility. Region availability and current details may vary by model and backend supply.
FAQ
Pricing questions before production.
Sensitive billing policy should be confirmed in the model page, console, docs, or billing decision notes before launch.
Is there a free tier?
Check the current console and account plan details. This public page does not invent promotional or free-tier policy.
Do I need separate provider accounts?
AllToken is designed to reduce provider-by-provider setup by giving teams one API and billing layer for supported routes.
How are failed requests billed?
Failed request treatment depends on backend settlement rules and task state. Confirm the current policy in billing documentation before production use.
How do video models differ from text pricing?
Video generation is async and may depend on model, duration, aspect ratio, resolution, generated output, and final settlement.
How are image generation tasks billed?
Image generation can follow task and generated-output semantics. Use current docs or console details as the source of truth.
Can I estimate cost before production?
Use model details, request patterns, and test traffic to estimate before scaling production usage.
Where can I see actual model pricing?
Use the model catalog, model detail pages, or authenticated console where available.
Why does this page not show every live model price?
Live prices, discounts, and provider routes can change. Use model pages and account usage records as the source of truth.
Start
Start with one API key.
Browse models, read the docs, then route production traffic through the AllToken account and billing surface when you are ready.