Put up to four models side by side: price, context, capabilities and modalities, straight from the catalog. Prices are what you pay through Rapid Router.
The priciest tier from the biggest labs, which is where the strongest models sit.
gpt-4 · claude-fable-5 · nemotron-4-340b-instruct
The cheapest live chat models that still cost something, for high-volume use.
llama-3.2-1b-instruct · qwen3.5-0.8b · glm-5.3-flash
The largest context windows in the catalog, for whole-repo and long-document work.
grok-4.20 · grok-4.20-non-reasoning · gpt-5.4
Models that reason before answering, for multi-step problems.
o1 · claude-fable-5 · gemini-4-argon
Pick at least two models, or start from a card above.