Is GLM free?
Some GLM models are listed as free for API use, but GLM is not universally free. Z.AI’s published API pricing lists GLM-4.5-Flash as Free across the displayed token-pricing categories. The same rate card charges for other models, including GLM-4.5. So the answer depends on the model and provider—not just the GLM name. (docs.z.ai)
The rate card says prices are in U.S. dollars. It does not specify a free-use quota, rate limits, or overage terms for GLM-4.5-Flash. “Free” on the rate card therefore tells you its listed token prices, but not whether your particular deployment will have unlimited access or meet production requirements. (docs.z.ai)
Z.AI GLM API prices
Z.AI’s rate card lists these prices per 1 million tokens:
| Model | Input | Cached input | Output |
|---|---|---|---|
| GLM-4.5-Flash | Free | Free | Free |
| GLM-4.5 | $0.60 | $0.11 | $2.20 |
| GLM-4.5-Air | $0.20 | $0.03 | $1.10 |
Prices are in USD. The rate card also shows a separate cached-input-storage column; check the official table for its terms and any other applicable charges. These are Z.AI rates, not prices for every provider offering GLM. (docs.z.ai)
Example: estimate a GLM-4.5 API request
Suppose a workload uses 1 million uncached input tokens and 100,000 output tokens with GLM-4.5. Using the listed rates:
- Input: 1 × $0.60 = $0.60
- Output: 0.1 × $2.20 = $0.22
- Estimated total: $0.82
This is an illustrative calculation, not a tested bill. It assumes those token volumes, uncached input, and the listed rates; actual usage and any applicable pricing terms can change the total. (docs.z.ai)
Is the free model ready for a production API?
A free listed token price is not the same as a production guarantee. Before deploying GLM-4.5-Flash, confirm that it is available for your intended API route and check the applicable usage limits, availability, support, and any service commitments. The pricing page establishes the listed rates, but does not provide those operational terms. (docs.z.ai)
Z.AI documents a standard API endpoint, bearer-key authentication, and a chat-completions example. Its GLM-4.5 overview describes features such as streaming, function calling, context caching, and structured output for the series. Those docs do not, by themselves, confirm every feature or operational term for GLM-4.5-Flash. (docs.z.ai)
Bottom line: If you mean Z.AI’s hosted API, GLM-4.5-Flash is listed as free on its rate card. Other GLM models have token charges, and the published free listing does not spell out quotas or production service guarantees. Check the rate card and model-specific API documentation for the exact model and route you plan to use.
Sources
Pricing - Z.AI Developer Documentation