GLM and ChatGLM API Pricing Guide
GLM models appear through Zhipu AI, Z.AI and international aggregators. A low listed rate does not establish the total bill, and “free” variants can have different limits or availability. This guide keeps those channels separate and calculates a consistent workload example.
Current API price snapshot
Up to 12 tracked language models are shown. Channels are selected by lowest input rate normalized to USD, not lowest total cost. Examples use 1M input plus 250K output tokens, without additional cache, batch or context-tier adjustments or tool fees. Processing-mode variants are not interchangeable; no total is shown when output pricing is unknown.
| Model | Lowest-input-rate channel | Input / 1M | Output / 1M | Example cost (USD) | Channels | Selected price verified |
|---|---|---|---|---|---|---|
| GLM 5.3 Prime | OpenRouter | $2.80 | $8.80 | $5.00 | 1 | Oct 17, 2026 |
| GLM 5.3 Flashx | OpenRouter | $0.37 | $1.25 | $0.6825 | 1 | Oct 17, 2026 |
| GLM Flash | OpenRouter | $0.06 | $1.60 | $0.46 | 1 | Oct 17, 2026 |
| GLM 5.3 Flash Batch | OpenRouter | $0.05 | $0.18 | $0.095 | 1 | Oct 17, 2026 |
| GLM 5.3 Flash | XycAi | ¥0.39 | ¥0.98 | $0.0923 | 5 | Oct 17, 2026 |
| GLM | OpenRouter | $0.06 | $4.40 | $1.1632 | 1 | Oct 17, 2026 |
| GLM 5.3 Batch | OpenRouter | $0.45 | $2.00 | $0.95 | 1 | Oct 17, 2026 |
| GLM 5.3 | OpenRouter | $0.08 | $4.80 | $1.279 | 5 | Oct 17, 2026 |
| GLM 5.2 | XycAi | ¥3.92 | ¥13.72 | $1.0652 | 5 | Oct 17, 2026 |
| GLM 5.1 | OpenRouter | $0.97 | $3.04 | $1.725 | 2 | Oct 17, 2026 |
| GLM 5v Turbo | OpenRouter | $1.20 | $4.00 | $2.20 | 1 | Oct 17, 2026 |
| GLM 5 Turbo | OpenRouter | $1.20 | $4.00 | $2.20 | 1 | Oct 17, 2026 |
Official versus free endpoints
Use the official baseline when support, account ownership, and predictable quotas matter. Aggregator free routes are useful for evaluation, but they may apply separate rate limits, routing rules, or availability windows.
Mainland China access
Zhipu’s China service is generally the direct-access option for mainland users. International channels can differ in payment methods, data location, model naming, and release timing, so compare the exact channel row before integrating.
Cost formula: keep CNY and USD ledgers separate
For metered GLM calls, cost = ordinary input ÷ 1M × input rate + cache-hit input ÷ 1M × cached rate + output ÷ 1M × output rate + applicable cache storage and tool charges. Use CNY rates on the China BigModel site and USD rates on Z.AI; convert only after calculating each bill. Resource packs are prepaid purchases with model coverage and validity conditions: allocate their purchase cost to covered usage, then add uncovered metered usage, without charging the covered tokens twice. A GLM Coding Plan has its own endpoint and quota rules; its subscription price is not a general API token allowance.
Usage example: a paid route and a Flash evaluation
Assume 2,000 paid GLM requests averaging 2,000 input and 500 output tokens, with no cache hits, tools or resource pack. The illustration uses 4M input and 1M output: cost = 4 × I + O, where I and O are the exact model’s live rates in the chosen site’s currency. A separate evaluation using a variant explicitly listed as free can have zero model-token charges within its limits; do not carry that assumption over to every model containing “Flash”. If a resource pack covers some requests, calculate its allocated cost and the uncovered requests separately.
Easy-to-miss costs: packs, tools and free-route limits
Check resource-pack expiry and supported models before prepaying; unused quota changes your effective cost. A free Flash variant still has rate and concurrency limits, and a paid fallback creates a separate bill. Tool calls and cache storage have their own pricing entries. Coding Plan quotas do not remove ordinary API charges, and a currency conversion on this page does not change the currency charged by either official site.
How to use this data
- Recalculate with your real input/output token mix.
- Check region, rate limits, and payment method.
- Recheck official sources before purchasing or switching.
Frequently asked questions
Are BigModel and Z.AI prices in the same currency?
The China site lists CNY prices; Z.AI lists USD prices. Match the exact model, token unit and endpoint before converting the resulting totals.
Is every GLM Flash model free?
No. Use only variants explicitly marked free in the official table, and check their limits. Flash and FlashX names alone do not establish a zero rate.
Does a Coding Plan cover ordinary API calls?
Treat the coding subscription and metered API as separate products. Check the coding endpoint and allowed usage rather than deducting its monthly price from an API bill.
How should I budget for a resource pack?
Check coverage and expiry, allocate the prepaid cost across usage you expect to consume, then add charges outside the pack. Unused capacity is still part of your purchase cost.
Official sources and review date
Billing guidance reviewed: . Price verification dates are listed separately in the table.