GLM-5.3 Uncensored goes free: 1.05M context, 131K output, 108 tok/s

petrusenko_max · x · 2026-10-25

Per @Rivonn, GLM-5.3 Uncensored — an abliterated research variant with the refusal direction removed from the weights — is now free via an OpenAI-compatible API. Specs: 1.05M context, 131K max output, 108 tok/s; the screenshot shows the usual $1.40/$4.40 per million token pricing zeroed out during the free promotion.

Original post →

More from Models

Models channel →