GLM 4.7 API Drops 10%: Input Now $0.54, Output $1.98 per 1M Tokens

What Changed

Z.ai has trimmed GLM 4.7 API pricing by -10% on both input and output tokens. Input drops from $0.60 to $0.54 per 1M tokens; output moves from $2.20 to $1.98 per 1M tokens. No capability changes are attached to this update — it's a straight cost reduction.

Does It Move the Needle?

A 10% cut is modest but real, especially on output-heavy workloads. Output tokens are where most production costs accumulate — long completions, RAG pipelines, chat sessions with verbose responses. If your app generates significant output volume, the drop from $2.20 to $1.98 per 1M tokens compounds quickly at scale.

For input-dominant use cases — batch classification, embedding-adjacent tasks, short-prompt workflows — the saving is smaller in absolute terms: $0.06 per 1M tokens. Still free money, but nothing to restructure a budget around.

Context

GLM 4.7 sits in a competitive mid-tier alongside several models in a similar price band. Whether this cut makes it the right pick depends on your quality-per-dollar benchmarks for your specific task, not the price reduction alone. Check GLM 4.7 — live specs & price history to stack it against alternatives currently tracked on the site.

Bottom Line

If you're already running GLM 4.7, your costs just fell 10% with no migration work required. If you've been evaluating it against pricier options, the new output rate of $1.98 per 1M tokens makes the comparison slightly more favorable. Worth a re-run of your cost model.

Ezra, Scout AI Team