GLM 5.2 API Prices Jump 245%: What It Costs You Now

The Numbers

Z.ai has repriced GLM 5.2 significantly. Input tokens move from $0.28 to $0.966 per million, and output tokens from $0.88 to $3.04 per million — a +245% increase across the board.

There's no soft way to frame that: it's a steep jump, not a minor adjustment.

Does It Hit Your Budget?

The impact depends almost entirely on your token mix and volume:

• Low-volume or input-heavy workloads (think summarization or classification pipelines) will feel the output price rise less, but input cost is still climbing by roughly 3.5×.
• High-output tasks — code generation, long-form drafting, agentic loops — take the hardest hit, with output now at $3.04/M.
• At scale, a workload burning 100M output tokens monthly goes from $88 to $304. That's real money.

What to Do

Before absorbing the increase, it's worth benchmarking GLM 5.2 against models in a similar price tier. Several alternatives on this site now undercut its new output price while matching or exceeding it on standard reasoning and language tasks.

If you're mid-contract or have budget locked in at the old rate, check with Z.ai on transition timelines — pricing changes like this sometimes include a grace window for existing integrations.

For live specs, the updated price tiers, and side-by-side comparisons, see GLM 5.2 — live specs & price history.

Bottom Line

GLM 5.2 remains a capable model, but at $3.04/M output it's no longer cheap-tier. Reassess if cost efficiency was your primary reason for choosing it.

Ezra, Scout AI Team