Find your model | AI Models Info

Step 1 of 6 What are you building?

This sets which qualities matter most in the ranking.

Writing and reviewing code Building features, fixing bugs, refactoring, code review Autonomous agents Multi-step tasks with tool use, running with little supervision Research and analysis Hard reasoning, scientific and technical questions, deep analysis Writing and content Articles, marketing copy, documentation, editing Chat and support Customer-facing assistants where latency and cost matter most Classification and extraction Tagging, routing, pulling structured data out of text, at volume Images and documents Reading screenshots, charts, scanned documents and diagrams Long documents Whole codebases, contracts, books, large transcripts

Step 2 of 6 What matters most?

Every model trades something off. Tell us what you are willing to trade.

Getting the best answer Accuracy matters more than cost or latency. A sensible balance Good enough quality without overpaying. Keeping costs down High volume, tight margins. Responding fast A user is waiting on the answer.

Step 3 of 6 Any budget ceiling? Optional

Blended cost per million tokens, at a 3:1 input-to-output ratio.

No ceiling Under $1 Under $5 Under $15

Step 4 of 6 How much context do you need? Optional

The size of a single request: documents, code, conversation history.

Standard is fine At least 128k tokens At least 500k tokens At least 1M tokens

Step 5 of 6 Any hard requirements? Optional

Models missing any of these will be excluded, not just ranked lower.

Read images Call tools or functions Extended reasoning Guaranteed structured output Prompt caching Batch processing

Step 6 of 6 Any hosting or data constraints? Optional

Relevant if you have compliance obligations or need to run models yourself.

No constraints EU data residency Inference must be available in an EU region. Must be self-hostable Run on your own hardware. Open weights Weights must be publicly downloadable.