Compare the tradeoffs.

Your models, your workload, and the evidence behind the choice.

Model library Find a provider Compare hosts

Your workloadChange the assumptions. See the economics.

Monthly requests

Input tokens / request

Output tokens / request

Optimize for

Evaluate

Cache hits · share of input

Cache writes · share of input

Extra volume from retries

Allowed gap from task leader

points

Minimum context capacity

tokens

Minimum task index

points

Cache rates apply only when supplied for the exact endpoint. Otherwise the standard input rate is used and the estimate is flagged. Extra reasoning tokens and provider-specific fees are excluded. Context fit checks the greater of input plus output tokens and your minimum capacity.

Restoring your workspace…

Know what you’re comparing.

Monthly cost is a token-only projection. Response time is an approximation from source performance measurements. Cache rates are endpoint-specific when available; model-level cache reference prices are not assumed to apply to other hosts. Missing results and methodology versions remain visible.

Using saved data exported 11 Nov 2026, 10:48 UTC. Fetch dates are not measurement dates. Sources & methodology

Find your next step

WhatLLM

Keep your workspace.

Compare the tradeoffs.

Add an exact model variant

Save this decision

Share this decision