Plans

Plans in Tukun.ai control product capacity and usage. They are not just a pricing page concern. They affect how an account can onboard, who owns cost, and how broadly the team can operationalize the product.

The billing boundary

Billing belongs to the account, not to an individual user identity.

That matters because:

• usage should reflect shared operational reality
• one team may have multiple users but one billing owner
• plan changes should align with account-level rollout decisions

Credit pricing

For billing reference, 1,000 credits equals USD 1 of metered usage.

That means:

• included plan credits map to a clear dollar-equivalent usage budget
• different models can still consume credits at different rates
• stronger models usually consume credits faster than lighter models
• included monthly credits are consumed before any additional metered usage

Current plan rules

The current personal plans also carry a few runtime rules that affect day-to-day usage.

Model selection

Paid plans can use Auto model selection. Auto does not mean unlimited provider access. It means Tukun.ai chooses a concrete model inside the active plan for the current turn.

In normal use, Auto starts from the plan’s default model. It only moves to a stronger eligible model when the turn needs a specific capability, such as requested reasoning support or analysis-grade execution.

This keeps everyday questions predictable while still giving heavier analysis room to use a better-suited model. Because different models may consume credits at different rates, Auto can still affect how quickly the account uses its included credits.

For the current plan-by-plan model contract, see Models by plan. That page explains what each plan can expect from the model catalog, what Auto actually does, and which details are product contract versus runtime implementation.

Conversation turn limits

• Free: up to 24 turns per conversation
• Pro: up to 50 turns per conversation
• Max: up to 50 turns per conversation

These limits apply per conversation thread. If a workflow routinely needs longer multi-step analysis, evaluate the plan against that real usage pattern instead of only comparing monthly credits.

Overage

At the moment, Free, Pro, and Max do not support credit overage.

That means:

• usage stays within the credits included in the active plan
• when included credits are exhausted, the product does not continue by charging extra metered credits automatically
• if this policy changes later, the billing and pricing docs should be updated as the source of truth

How to think about plan evaluation

Do not evaluate plans in isolation. Evaluate them against the workflow you are trying to support.

Ask:

• how many people need to use the account regularly
• how many recurring questions will be operationalized
• how much analysis volume the team expects
• whether the product is still in one-owner onboarding or broader team use

A good evaluation pattern

During early evaluation:

• keep one clear business question in scope
• use one or two representative sources
• decide whether the product will become a personal workflow or a shared operating workflow

That gives you a real basis for deciding the right plan path.

Before a broader rollout

Before expanding usage, confirm:

• the billing owner is clear
• source ownership is clear
• the team understands current limits
• the key metrics are stable enough to be reused
• support and security contacts are known

When a plan change is justified

Consider a plan change when:

• more people need shared access
• usage has moved from exploration to regular operations
• the account now owns recurring dashboards or broader semantic assets
• the current capacity limits are blocking real team workflows

What not to do

Do not upgrade just because one demo went well. Upgrade when the product is becoming part of an actual operating loop.