Skip to main content
AI Model ComparatorAI Model Comparator

Settings

Manage model data sources

Ollama model data

The Ollama library catalog (context, parameter sizes, capabilities) plus live cloud pay-as-you-go pricing from ollama.com/pricing are used to mark each model as cloud-served, downloadable, or both.

Never refreshed manually (using seeded catalog)

Refreshing re-reads the Ollama library catalog and the live cloud pricing table, recomputes which models are cloud-served vs downloadable, and updates the cached catalog. Existing AI summaries are preserved. This takes a few seconds.

OpenRouter model data

The full OpenRouter model catalog — pricing, context window, capabilities and free-tier variants — is pulled live from the OpenRouter models API.

Never refreshed manually (using seeded catalog)

Refreshing re-reads the live OpenRouter models API (prices in USD per 1M tokens) and updates the cached catalog. Existing AI summaries are preserved. This takes a few seconds.

OpenAI model data

Flagship models, pricing, reasoning & speed ratings, knowledge cutoff and promotional pricing are pulled from OpenAI’s developer docs.

Never refreshed manually (using seeded catalog)

Refreshing re-reads OpenAI’s live pricing and per-model pages, then updates the cached catalog used across the comparison view. Existing AI summaries are preserved. This takes a few seconds.

SpaceXAI (Grok) model data

Grok models, pricing, context window and capabilities (vision, tools, reasoning) are pulled from the SpaceXAI (x.ai) developer docs.

Never refreshed manually (using seeded catalog)

Refreshing re-reads the SpaceXAI live pricing catalog and per-model pages, then updates the cached catalog. Existing AI summaries are preserved. This takes a few seconds.

Anthropic (Claude) model data

Claude models, pricing, context window, release date, knowledge cutoff and capabilities (vision, tools, reasoning) are pulled from the live models.dev catalog.

Never refreshed manually (using seeded catalog)

Refreshing re-reads the live Anthropic catalog (prices in USD per 1M tokens) and updates the cached catalog. Existing AI summaries are preserved. This takes a few seconds.

Google Gemini model data

Gemini models, pricing, context window, release date, knowledge cutoff and capabilities (vision, tools, reasoning) are pulled from the live models.dev catalog.

Never refreshed manually (using seeded catalog)

Refreshing re-reads the live Gemini catalog (prices in USD per 1M tokens) and updates the cached catalog. Existing AI summaries are preserved. This takes a few seconds.