Compare AI models
These are the model comparisons GPTProto has published. Each page stays with two named models and shows the listed price, the context window or output size, and the workloads worth trying first. Speed and benchmark figures appear only when that page cites a source and a date. Open a comparison, then confirm the live rate on the model page before you budget.
Compare models from the same provider
These matchups stay inside one provider, so the price unit and the product role are comparable. This index lists only pages that exist.
Anthropic
- Claude Haiku 5.5 vs Claude Opus 5.5↗Price, speed, context and the workloads each model fits.
- Claude Haiku 5.5 vs Claude Sonnet 5.5↗Price, the 100K-token tier, speed, and which bounded tasks belong on each model.
- Claude Opus 5.5 vs Claude Sonnet 5.5↗Half the token price on Sonnet, and where Opus still earns the difference.
OpenAI
Looking for one model
A comparison is a starting point. Open the catalog for current pricing, API examples, and a Playground on one model. Open pricing when you want to see how platform credit works before you call an API.
Questionsbefore you open a pair
These answers describe this index. Each comparison page has its own questions.
01What does a comparison page include?
Each published page covers two named models. It shows listed GPTProto and official prices where the catalog has them, plus context or output size and which workloads to try first. Latency and benchmark rows appear only with a cited source and date. They are not a GPTProto service level.
02Does this index list every model pair?
No. It lists only comparisons GPTProto has published. It does not create a page for every pair in the catalog.
03Are the prices on a comparison the live rate?
The comparison records a catalog snapshot and links to both model pages. Check those pages for the rate, route, and settings you plan to use.