Side-by-side cost instrument

Compare AI model costs

Run one workload through up to four models. TokenMeter applies each provider's current cache and context-band rules before ranking the result.

Comparison workloadUp to 4 models

Set your workload

1/4 models

Workload presetChoose an example starting point or enter your measured workload directly.

Monthly input tokensMonthly output tokensCached input tokensAverage input tokens per requestUsed only where providers charge a long-context rate.

Live monthly estimate

calibrated

Anthropic

Claude Fable 5.1

Lowest cost

$100.00/ month

Input: $44.44
Output: $55.56
Cached: $0.00000000
Context rate: Flat

Rate alternativessource-backed only

Cache-readSet cached input tokens above zero to estimate cache-read savings.

Unavailable

BatchBatch pricing is asynchronous and can have separate completion windows, eligibility, and limits; compare it only when that delivery model fits the workload.

Save $50.0050% · $50.00 optimized

Verified Oct 5, 2026Official pricing ↗