Z.aiopen weightsGLM 5.2 pricing
$0.62 per 1M input tokens and $2.42 per 1M output tokens. In real work that is 0.43¢ per chat answer and 1.2¢ to summarize a 20-page PDF.
Official price at Zhipu (Z.ai): $1.40 / $4.40 per 1M, cached $0.26 (source, verified Oct 18, 2026). OpenRouter lists $0.62 / $2.42, which is the cheapest route across its hosts - lower than the official price.
What GLM 5.2 costs per task
Token counts estimated from a measured model of the same lab (GLM 5.3 Flash), rescaled by tokenizer. Reasoning tokens are billed as output.
| Task | Cost | Tokens used | Price rank | Basis |
|---|---|---|---|---|
| Answer a chat question | 0.43¢$4.34 / 1,000 | 48 in · 391 out · 1,391 reasoning | 218 of 336 | estimated |
| Reply to a customer email | 0.08¢$0.79 / 1,000 | 265 in · 234 out · 24 reasoning | 148 of 336 | estimated |
| Write a 1,500-word blog post | 1.7¢$16.79 / 1,000 | 62 in · 2,174 out · 4,749 reasoning | 227 of 336 | estimated |
| Write 10 product descriptions | 1.7¢$17.22 / 1,000 | 289 in · 1,101 out · 5,940 reasoning | 262 of 336 | estimated |
| Translate 1,000 words | 2.5¢$25.40 / 1,000 | 1,330 in · 1,394 out · 8,761 reasoning | 264 of 336 | estimated |
| Proofread one page (500 words) | 0.24¢$2.38 / 1,000 | 730 in · 642 out · 155 reasoning | 167 of 336 | estimated |
| Summarize a 20-page PDF | 1.2¢$12.45 / 1,000 | 12,995 in · 544 out · 1,270 reasoning | 208 of 336 | estimated |
| Extract data from an invoice | 0.27¢$2.74 / 1,000 | 391 in · 455 out · 576 reasoning | 187 of 336 | estimated |
| Analyze a 200-row spreadsheet | ≥4.2¢$41.76 / 1,000 | 4,900 in · 0 out · 16,000 reasoning | 223 of 336 | estimated |
| Answer from 5 retrieved documents (RAG) | 0.71¢$7.07 / 1,000 | 3,714 in · 274 out · 1,697 reasoning | 208 of 336 | estimated |
| Classify 100 support tickets | 2.0¢$19.80 / 1,000 | 1,281 in · 441 out · 7,411 reasoning | 259 of 336 | estimated |
| Review 500 lines of code | ≥4.2¢$41.53 / 1,000 | 4,532 in · 2 out · 15,998 reasoning | 244 of 336 | estimated |
| Write a SQL query | 0.39¢$3.94 / 1,000 | 128 in · 473 out · 1,123 reasoning | 203 of 336 | estimated |
| Run a 50-step coding agent | 46¢$459 / 1,000 | 2,239,737 in · 15,014 out · 28,800 reasoning | 162 of 336 | modeled |
Price at every provider
List prices per 1M tokens (input / output) at each host serving GLM 5.2 through OpenRouter today. Hosts can differ in quantization, context and speed.
| Provider | Input / output | Cached input | Context | Uptime |
|---|---|---|---|---|
| B Baidufp8 | $0.19 / $0.59 | $0.035 | 1M | 100% |
| I Inceptronfp4 | $0.69 / $1.04 | $0.14 | 1M | 100% |
 DeepInfrafp4 | $0.45 / $1.44 | $0.084 | 1M | |
| S StreamLakefp8 | $0.56 / $1.75 | $0.10 | 1M | 100% |
DeepInfrafp4 | $0.56 / $1.80 | $0.10 | 1M | 100% |
 Novitafp8 | $0.65 / $2.04 | $0.12 | 1M | 100% |
| D DigitalOceanunknown | $0.70 / $2.20 | $0.10 | 1M | 100% |
 SiliconFlowfp8 | $0.70 / $2.20 | $0.13 | 1M | 100% |
| C CoreWeavefp4 | $0.62 / $2.42 | $0.14 | 1M | 100% |
| A AtlasCloudfp8 | $0.94 / $2.95 | $0.17 | 1M | |
 Alibabafp8 | $0.97 / $3.04 | $0.19 | 1M | |
| P Phalafp8 | $1.26 / $3.00 | $0.22 | 1M | |
Alibabafp8 | $1.12 / $3.52 | $0.22 | 1M | |
 Nebiusfp4 | $1.40 / $4.40 | - | 1M | 100% |
 Mistralnvfp4 | $1.40 / $4.40 | $0.14 | 1M | 100% |
| B BaseTenfp8 | $1.40 / $4.40 | $0.14 | 1M | 100% |
 Togetherunknown | $1.40 / $4.40 | $0.26 | 1M | 100% |
| B BaseTenfp8 | $1.40 / $4.40 | $0.14 | 1M | 100% |
| V Venicefp8 | $1.40 / $4.40 | $0.26 | 1M | |
| G GMICloudfp8 | $1.40 / $4.40 | $0.26 | 1M | |
 Parasailfp4 | $1.40 / $4.40 | $0.26 | 262K | 100% |
| F Friendliunknown | $1.40 / $4.40 | $0.26 | 1M | 100% |
Z.AIfp8 | $1.40 / $4.40 | $0.26 | 1M | 100% |
Mistralunknown | $1.54 / $4.84 | $0.15 | 1M | |
| R Relaceunknown | $0.084 / $6.00 | $0.084 | 1M | |
| M Morphfp8 | $0.088 / $6.00 | $0.055 | 1M | 100% |
| W Waferfp4 | $0.034 / $7.00 | $0.033 | 1M | 100% |
| B BaseTenfp8 | $2.10 / $6.60 | $0.21 | 1M | 100% |
| B BaseTenfp8 | $2.10 / $6.60 | $0.21 | 1M | 100% |
Alibabafp8 | $2.31 / $7.26 | $0.46 | 1M | 0% |
| C Cloudflareunknown | $0.19 / $8.00 | $0.19 | 262K | 100% |
| B Baidufp4 | $2.25 / $7.88 | $0.56 | 1M | 100% |
| D Decartfp4 | $2.50 / $9.00 | $0.54 | 1M | 100% |
| W Waferfp4 | $0.19 / $10.00 | $0.18 | 1M | 100% |
Price history
First seen in our price history on Jun 27, 2026 at $0.95 input / $3.00 output. Since then the output price moved from $3.00 to $2.42 per 1M tokens (12 changes).
FAQ
How much does GLM 5.2 cost?
GLM 5.2 costs $0.62 per 1 million input tokens and $2.42 per 1 million output tokens, with cached input at $0.14 (list price on Oct 18, 2026). In practice one chat answer costs about 0.43¢ and summarizing a 20-page PDF about 1.2¢.
Where is GLM 5.2 cheapest?
Among 34 providers we track, the lowest list price today is at Baidu: $0.19 input and $0.59 output per 1M tokens.
Does GLM 5.2 charge for reasoning tokens?
Yes. Reasoning (thinking) tokens are billed as output tokens. That is why the cost per task can be much higher than the visible answer suggests - our task costs include them.
Is the GLM 5.2 cost per task measured or estimated?
Estimated: GLM 5.2 was not run directly. Tokens come from GLM 5.3 Flash and are rescaled by the tokenizer of Z.ai. Real usage can differ.