Normalized pricing · Usage-based monthly estimate · Share + export decision brief

Find the cheapest model that still works.
Estimate monthly cost, not just unit price.

Filter by task/capabilities, estimate monthly spend from your input/output token assumptions, and export a Markdown/CSV brief for reviews and approvals.

Refine in Models Compare finalists

Filters & usage assumptions

Task
Requirements
Monthly input (1M tokens)
Monthly output (1M tokens)
Min context (tokens)
Baseline (for savings)
—
Candidates
Matches filters
—
Providers
Coverage
—
Lowest input price
$/1M tokens
—
Lowest monthly cost
Under assumptions
Rank Model Provider Input Output Monthly Context
Savings (vs baseline)
—
Select a model to see savings
Selected (monthly)
—
Baseline (monthly)
—
Validate before buying

Don't buy on price alone

Test finalists with your real prompts in Playground. Then export a brief instead of screenshots so decisions stay reproducible.

Open Playground View rankings
Methodology

Why "cheapest" here is more defendable

We normalize pricing to $/1M tokens and include both input + output in the same monthly cost formula to avoid hidden "cheap input, expensive output" traps.

Usage-based estimates
Input and output pricing differ. We calculate monthly cost directly so you don't get surprised by output fees.
Open Pricing →
Compare + export
Export Markdown/CSV and share links for reviews, approvals, and documentation.
Open Compare →
Transparency
Clear advertising + data handling disclosures reduce policy and trust risks.
See policy →