Claude Haiku 5.5 (Xhigh) vs GPT-6 Astra (Max)

What the captured data shows

Haiku 5.5 Xhigh has lower listed token rates and a lower intelligence index than GPT-6 Astra Max. Haiku’s captured timing values show a shorter first-token delay and higher output throughput. Both entries have no individual benchmark results in the current records, limiting task-specific conclusions.

Which model fits your workflow

Run difficult examples that make the Max Astra configuration valuable to you. Inspect whether Haiku Xhigh satisfies the same requirements, and include the cost of any repair attempt. Preserve the exact settings and record completion time rather than judging by streamed output speed alone.

Decision takeaway

Haiku Xhigh offers a lower-priced configuration with favorable recorded timing; Astra Max has the stronger captured aggregate index.

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 (Xhigh) and GPT-6 Astra (Max). Missing measurements remain unavailable; this page does not report a new hands-on test.