01
Match capability
Use a general chat model for drafting and research. Choose a reasoning model only when the task benefits from deeper step-by-step work.
Tavory blog
A calmer way to follow the AI landscape.
Practical notes about choosing models, staying in control of costs and building a trustworthy AI workflow. The live catalog signal below updates with Tavory's available models.
Explore live models Live catalog signal
Refreshing0+
models
0+
providers
0+
multimodal
The catalog refreshes automatically as providers update their availability.
From the journal
Useful, not noisy.
We do not automatically republish third-party news or images. Each Tavory article links to a product area or original provider information, so the page stays reliable.
Guide
Choosing a model without losing track of cost
A practical guide to model capability, context size and provider pricing inside a single workspace.
Read moreProduct
What Smart and Saver mode actually optimize
How Tavory balances quality, speed and cost while keeping the final model choice visible.
Read moreSecurity
Privacy when working across AI providers
The controls that matter when prompts, files and provider access need to remain understandable.
Read moreDaily AI briefing
The signal behind the release cycle.
Tavory turns licensed benchmark data into short, original daily notes. Rankings are contextual: intelligence, speed and token price are different measurements.
Checking the latest published briefing.
A working method
Start with the task, not the brand.
The best model is rarely the one with the loudest launch. Start with the type of work, then decide how much reasoning, context and multimodal capability it actually needs.
02
Match the input
For images and documents, select a model that explicitly supports vision or files. A text-only model cannot inspect an attachment.
03
Match the budget
Use Saver for routine work, Smart for balanced choices, or select a model manually when a project has a specific quality or cost target.
Credits explained
A budget should make decisions easier.
Tavory separates access from usage. A plan determines which model range and limits are available. Your AI budget tracks the provider cost of the work you choose to run.
Compare plansFree
Start with no-cost models. Optional Flex credits add usage for economical Flex models, but do not turn a Free account into a Premium-model account.
Pro
For individual work that needs access to Premium models, a higher request limit and a monthly included AI budget.
Team
For shared work where projects, capacity and cost control need to be visible across a team.
Latest catalog changes
Models are moving targets.
Availability and provider pricing change constantly. Tavory refreshes its catalog so model selection is based on what is currently offered, not on a static list.
The live catalog is loading.
A fair comparison
Tavory complements providers. It does not replace their models.
Tavory is the workspace layer around providers: it gives you choice, context and cost visibility while the underlying models remain provided by their respective companies.
| Workflow | Tavory | Provider chat | API console |
|---|---|---|---|
| Model choice | Multiple providers in one catalog | One provider at a time | You wire providers yourself |
| Cost visibility | Request costs and AI budget in CHF | Provider-specific billing | Raw API usage and invoices |
| Projects and memory | Built into the workspace | Varies by provider | Requires your own implementation |
| Files and tools | One workspace flow | Varies by provider | Requires your own implementation |
Use Tavory when you want to compare and work across providers; use a direct provider product when you specifically need that provider's proprietary experience.
Provider guide
The right comparison is about the task.
OpenAI, Claude, Gemini, Perplexity and the wider model ecosystem each have different strengths. Tavory is built for the moment when you want to compare those strengths, switch providers and keep the work in one calm place.
OP
Provider viewOpenAI
General chat, coding and multimodal work
A broad ecosystem with capable general-purpose models.
Why Tavory helps
Compare OpenAI models with other providers, projects, files and visible usage in one workspace.
AN
Provider viewAnthropic / Claude
Writing, analysis and long-context work
A strong fit for careful writing and complex document tasks.
Why Tavory helps
Keep Claude available alongside alternatives instead of making one provider your only workflow.
GO
Provider viewGoogle / Gemini
Multimodal tasks and large context
Useful when images, documents or large context windows matter.
Why Tavory helps
Choose by capability and current availability, then keep the result with the rest of your work.
PE
Provider viewPerplexity
Web-grounded research
Designed around research answers with web citations.
Why Tavory helps
Use web-ready models when needed and keep research, files and follow-up work together.
OT
Provider viewOther providers
Specialized, open and emerging models
Different providers can be better for speed, price, code or a specific modality.
Why Tavory helps
Tavory makes those trade-offs visible through one catalog, routing modes and cost controls.
Common questions
Before you start a new workflow.
The most useful AI setup is usually the one whose limits are visible: which provider is used, what a request costs and what happens to files.
Does Tavory build its own AI models?
No. Tavory is a workspace for working across providers. The selected provider remains responsible for the underlying model and its output.
When should I use Smart mode?
Use Smart mode when you want Tavory to select from the models available to your plan. It is useful for everyday work where you do not want to choose manually each time.
What is the difference between Flex credits and Pro?
Flex credits are one-off AI budget for free and economical Flex models. Pro unlocks Premium models, higher limits and an included monthly AI budget.
Can I use my own provider key?
Yes. Tavory supports a personal Requesty key in Settings. Requests charged to your own key use that provider account instead of Tavory AI budget.
Why can model availability change?
Providers can add, retire, reprice or limit models. Tavory refreshes its catalog so the model picker reflects current availability.