GLM 5.3 Input Price Rises 40% — What It Means for Your API Budget

What Changed

Z.ai has quietly pushed through a +40% increase on GLM 5.3 input token pricing. The rate moves from $0.05 to $0.07 per 1M tokens. Output pricing stays flat at $7.00 per 1M tokens.

Does It Actually Matter for Your Workload?

Depends on your input-to-output ratio. For most text generation tasks, output tokens dominate spend — so if you're pumping out long completions, this increase barely moves your total bill. But input-heavy workflows — retrieval-augmented generation, long-context summarisation, document processing — will feel it more directly. A pipeline feeding 10M input tokens a day just got $0.20 more expensive per day, or roughly $73 more per year, on input costs alone.

That said, $0.07/1M input is still on the cheaper end of the market. The question isn't whether the increase stings in absolute terms — it's whether GLM 5.3's price-to-performance ratio still holds up against alternatives now that the input cost has shifted.

What to Do Next

• Re-run your cost estimates if you last benchmarked GLM 5.3 at the old $0.05 rate.
• Check your input volume — if it's low relative to output, this change is mostly noise.
• Compare alternatives using current pricing. See GLM 5.3 — live specs & price history for up-to-date figures alongside competing models.

No explanation from Z.ai has been published for the increase. Whether this is a one-off correction or signals further adjustments is unclear.

Ezra, Scout AI Team