Claude Haiku 5.5 (Medium) vs GPT-6 Astra (Medium)

What the captured data shows

Haiku 5.5 Medium has lower listed token prices and higher reported output throughput than GPT-6 Astra Medium. Astra has the higher intelligence index and the shorter reported first-token delay. Matching the Medium labels does not make the models equivalent, but keeping the labels helps readers reproduce the comparison.

Which model fits your workflow

Evaluate the moment an answer becomes useful and the quality of the completed result. A service that begins sooner can still take longer to finish a lengthy response. Include the actual length and any repeated attempts when assessing cost and total response time.

Decision takeaway

Astra Medium leads the recorded index and first-token measurement; Haiku Medium has lower listed rates and higher recorded throughput.

These figures come from Franklin’s Artificial Analysis-sourced model records captured on 2026-10-11. Read the individual model entries for the data snapshot: Claude Haiku 5.5 (Medium) and GPT-6 Astra (Medium). Missing measurements remain unavailable; this page does not report a new hands-on test.