How GLM API pricing works
Z.ai, the company behind the GLM models, bills its API per token with separate input, cached-input and output prices. The current GLM-5.3 line has three options:
- GLM-5.3 ($1.40 / $4.40): Z.ai's flagship, built for coding and agentic work. It takes text only, always reasons before answering (with three effort levels), and has a 1M-token context window with up to 128K tokens of output.
- GLM-5.3 Flash ($0.15 / $0.50): the low-cost model, and the first natively multimodal model in the GLM-5 series, so it accepts images, video and files as well as text.
- GLM-5.3 FlashX ($0.37 / $1.25): the same model as Flash on faster serving, which Z.ai quotes at up to 200 tokens per second, for about two and a half times the price.
Cached input
Repeated input served from cache costs $0.26 per 1M tokens on GLM-5.3, $0.03 on Flash and $0.075 on FlashX, about a fifth of the normal input price. Z.ai also lists a storage charge for cached input but currently waives it as a limited-time offer. For a coding agent with 750M input tokens a month, 85% of it cached, and 8M output, GLM-5.3 costs about $358 a month and GLM-5.3 Flash about $40.00.
Z.ai doesn't list a batch discount on its pricing page.
Older GLM models
Earlier GLM models stay on sale. GLM-5.2 and GLM-5.1 cost the same as GLM-5.3 ($1.40 / $4.40), so there's little reason to stay on them for price. GLM-5 costs $1.00 / $3.20, and GLM-4.7 and GLM-4.6 cost $0.60 / $2.20. GLM-4.7-Flash is free to use.
GLM Coding Plan or the API?
Z.ai also sells the GLM Coding Plan, a subscription for using GLM models in coding tools, with a points-based quota instead of per-token billing. GLM-5.3 and GLM-5.3 Flash are available on it. Like other coding subscriptions, it suits one developer working in a coding agent every day, while the API suits apps and automation. See Z.ai's Coding Plan page for the current tiers and prices.
Which GLM model should you use?
- Coding and agents: GLM-5.3. Its input and output prices are well below Claude Sonnet 5.5’s and GPT-6.1 Sol’s, but its cached input ($0.26) costs more than theirs ($0.10), so for cache-heavy agents the monthly bill comes out about the same.
- High-volume work, and anything with images or video: GLM-5.3 Flash.
- Interactive apps where response speed matters: GLM-5.3 FlashX.
Other charges
- Web search as a built-in tool costs $0.01 per use.
- API formats: Z.ai accepts both OpenAI-compatible and Anthropic-compatible requests, so GLM works in tools built for either.
Sources: Z.ai pricing, GLM-5.3, GLM-5.3 Flash and FlashX. Checked 13 October 2026.
