Every AI model, in one place.

Pricing, benchmarks, provider latency, and how teams actually use each one.

86 matches

Large language models handle the core text work in most AI workflows: drafting content, answering questions, summarizing documents, generating code. Teams pick an LLM when they need flexible reasoning over natural language or structured data, not a narrow task-specific model. The defining technical attribute is context window size—models that hold 128k+ tokens let you process entire codebases or long transcripts in one pass, while smaller windows force chunking strategies that lose coherence. Choosing inside this category means balancing cost against output quality. Frontier models deliver stronger reasoning and fewer hallucinations but cost 10–50× more per token than mid-tier alternatives. Speed matters for user-facing features; latency varies 3–10× across models at similar capability levels. If your use case tolerates occasional errors and you're processing high volume, start cheaper and test whether accuracy gaps actually hurt your workflow before paying for top-tier inference.

Prefer one page? Browse all AI models A–Z.

About llm models