Reasoning Models in 2026
Updated
For tasks that call for reasoning, compare the models marked for that capability alongside their context capacity and fit for your application. This catalogue contains 332 entries and orders them by context, highest first, with filtering and sorting available. Grok 4.20, Grok 4.20 Multi-Agent and GPT-5.4 lead the displayed results; GPT-5.4 Pro and GPT-5.5 also appear near the start. Use context as the ordering signal, then weigh the details of each model against the requirements of the system you already run.
332 on record. Sorted by context; every figure comes from the OpenRouter model catalogue, with the page it was read from on each entry.
332listed
