mf Model Finder

MODEL FINDER · BY INDIGLOO

Find your fit.
Skip the hype.

The most expensive AI model isn’t always the right one. Compare your options and discover what meets your quality bar.

Explore across providersOpenAI · Anthropic · Google · open-weight models
THE DECISION SPACE01 / EXPLORE
Quality ↗Cost ↓Speed ↗
Your
best fit.DEFINED BY YOUR TASK
No universal winner.
Just a better match.

NEW / BROWSER ANALYTICS

See your usage.
Size your next model.

Import a usage JSON export to explore requests, tokens, sessions, activity, and reported costs. Estimate model-weight memory from parameters and precision.

Open analytics ↗ Plugin download and plans ↗

A LITTLE STRUCTURE. A BETTER DECISION.

From “which model?”
to a shortlist you can test.

01

Define the job

Describe the input, expected output, budget, privacy needs, and what a good answer looks like.

02

Compare the tradeoffs

Ask Model Finder to research suitable providers, current pricing, capabilities, and deployment options.

03

Test your quality bar

Use representative examples and a clear pass threshold. Choose the cheapest candidate that passes.

MAKE IT SPECIFIC

Your next
comparison starts here.

Build a brief to paste into Model Finder in ChatGPT. This website creates prompts; it doesn’t call model APIs or run benchmarks.

If Model Finder is installed, select it in ChatGPT before sending your prompt. Public marketplace availability is not confirmed.

CHECK THE MATH

What might it cost?

Enter your own provider prices. This estimate covers token charges only, excluding tools, storage, caching discounts, and taxes.

Enter both prices to estimate

Requests × (input tokens × input rate + output tokens × output rate) ÷ 1,000,000

EVIDENCE OVER LEADERBOARDS

Your examples are
the benchmark that matters.

Model Finder guides research and evaluation planning. Recommendations are task-specific, and provider prices and availability can change.

Start with your task ↗