PROVIDERS / ORIGINAL SOURCES

Follow the source.

Model publishers, hosted APIs and open-weight checkpoints each tell a different part of the story.

Qwen↗

Open-weight text and vision-language models. Compare the exact checkpoint, thinking mode and serving configuration.

General-purpose and reasoning model families. Distilled versions and dated revisions are distinct artifacts.

Llama open-weight models. Check the model license, modality support and total deployment requirements.

Hosted APIs and open-weight checkpoints. A model card and a provider API can describe different serving limits.

Gemini hosted APIs and Gemma open-weight models. Keep the product families, licenses and API formats separate.

Hosted model APIs with dated snapshots and distinct endpoint formats. Confirm current lifecycle and service terms.

Claude hosted models and the native Messages API. Compare tool loops, streaming and thinking controls on the actual endpoint.