Z.aiopen weightsreasoningGLM 5.3 Flash pricing
$0.048 per 1M input tokens and $0.20 per 1M output tokens. In real work that is 0.04¢ per chat answer and 0.10¢ to summarize a 20-page PDF.
Official price at Zhipu (Z.ai): $0.15 / $0.50 per 1M, cached $0.03 (source, verified Oct 18, 2026). OpenRouter lists $0.048 / $0.20, which is the cheapest route across its hosts - lower than the official price.
What GLM 5.3 Flash costs per task
Every row below was measured: we ran the task on this model and counted the tokens. Reasoning tokens are billed as output.
| Task | Cost | Tokens used | Price rank | Basis |
|---|---|---|---|---|
| Answer a chat question | 0.04¢$0.36 / 1,000 | 48 in · 391 out · 1,391 reasoning | 36 of 336 | measured |
| Reply to a customer email | <0.01¢$0.06 / 1,000 | 265 in · 234 out · 24 reasoning | 9 of 336 | measured |
| Write a 1,500-word blog post | 0.14¢$1.39 / 1,000 | 62 in · 2,174 out · 4,749 reasoning | 35 of 336 | measured |
| Write 10 product descriptions | 0.14¢$1.42 / 1,000 | 289 in · 1,101 out · 5,940 reasoning | 43 of 336 | measured |
| Translate 1,000 words | 0.21¢$2.09 / 1,000 | 1,330 in · 1,394 out · 8,761 reasoning | 105 of 336 | measured |
| Proofread one page (500 words) | 0.02¢$0.19 / 1,000 | 730 in · 642 out · 155 reasoning | 9 of 336 | measured |
| Summarize a 20-page PDF | 0.10¢$0.99 / 1,000 | 12,995 in · 544 out · 1,270 reasoning | 20 of 336 | measured |
| Extract data from an invoice | 0.02¢$0.23 / 1,000 | 391 in · 455 out · 576 reasoning | 11 of 336 | measured |
| Analyze a 200-row spreadsheet | ≥0.34¢$3.44 / 1,000 | 4,900 in · 0 out · 16,000 reasoning | 41 of 336 | measured |
| Answer from 5 retrieved documents (RAG) | 0.06¢$0.57 / 1,000 | 3,714 in · 274 out · 1,697 reasoning | 22 of 336 | measured |
| Classify 100 support tickets | 0.16¢$1.63 / 1,000 | 1,281 in · 441 out · 7,411 reasoning | 56 of 336 | measured |
| Review 500 lines of code | ≥0.34¢$3.42 / 1,000 | 4,532 in · 2 out · 15,998 reasoning | 45 of 336 | measured |
| Write a SQL query | 0.03¢$0.33 / 1,000 | 128 in · 473 out · 1,123 reasoning | 20 of 336 | measured |
| Run a 50-step coding agent | 2.6¢$25.63 / 1,000 | 2,239,737 in · 15,014 out · 28,800 reasoning | 8 of 336 | modeled |
Price at every provider
List prices per 1M tokens (input / output) at each host serving GLM 5.3 Flash through OpenRouter today. Hosts can differ in quantization, context and speed.
| Provider | Input / output | Cached input | Context | Uptime |
|---|---|---|---|---|
| O OpenInferencefp4 | $0.048 / $0.20 | $0.006 | 1M | 100% |
| S StreamLakefp8 | $0.075 / $0.25 | $0.015 | 1M | 100% |
 DeepInfrafp4 | $0.075 / $0.25 | $0.015 | 1M | 100% |
| D Decartfp4 | $0.083 / $0.28 | $0.017 | 1M | 100% |
 Novitafp8 | $0.084 / $0.28 | $0.017 | 1M | 100% |
| S Sail Researchfp4 | $0.078 / $0.30 | $0.052 | 1M | 100% |
| S Sail Researchfp4 | $0.078 / $0.30 | $0.052 | 1M | 100% |
| G GMICloudfp8 | $0.09 / $0.30 | $0.018 | 1M | 100% |
| I InferenceNetfp4 | $0.12 / $0.39 | $0.032 | 1M | 100% |
| P Phalanvfp4 | $0.13 / $0.42 | $0.025 | 1M | 100% |
| D DigitalOceanunknown | $0.13 / $0.44 | $0.03 | 1M | 100% |
| M Makoraunknown | $0.14 / $0.47 | $0.024 | 1M | 100% |
| R Relaceunknown | $0.075 / $0.50 | $0.007 | 1M | 100% |
| N Near AIfp8 | $0.08 / $0.50 | $0.012 | 1M | 100% |
| W Wafernvfp4 | $0.12 / $0.50 | $0.03 | 1M | 100% |
| M Modalnvfp4 | $0.15 / $0.50 | $0.03 | 1M | 100% |
| B BaseTenfp8 | $0.15 / $0.50 | $0.03 | 1M | 100% |
| C Crusoefp4 | $0.15 / $0.50 | $0.03 | 1M | 100% |
| C CoreWeavenvfp4 | $0.15 / $0.50 | $0.02 | 1M | 100% |
| A AtlasCloudfp8 | $0.15 / $0.50 | $0.03 | 1M | 100% |
 Fireworksunknown | $0.15 / $0.50 | $0.03 | 1M | 100% |
| F Friendliunknown | $0.15 / $0.50 | $0.03 | 1M | 100% |
 SiliconFlowfp8 | $0.15 / $0.50 | $0.03 | 1M | 100% |
 Togetherunknown | $0.15 / $0.50 | $0.03 | 1M | 100% |
 Parasailfp4 | $0.15 / $0.50 | $0.03 | 1M | 74% |
| B BaseTenfp8 | $0.15 / $0.50 | $0.03 | 1M | 100% |
| V Veniceunknown | $0.15 / $0.50 | $0.03 | 1M | 100% |
Z.AIfp8 | $0.15 / $0.50 | $0.03 | 1M | 100% |
| M Morphfp8 | $0.088 / $0.65 | $0.031 | 1M | 100% |
Parasailfp4 | $0.19 / $0.62 | $0.037 | 1M | 100% |
| I Inceptronfp8 | $0.10 / $0.69 | $0.075 | 1M | 100% |
| N NextBitfp4 | $0.25 / $0.70 | $0.075 | 1M | 100% |
Fireworksunknown | $0.23 / $0.75 | $0.045 | 1M | 100% |
| R Rekaunknown | $0.06 / $1.60 | $0.04 | 1M | 100% |
Price history
First seen in our price history on Sep 2, 2026 at $0.075 input / $0.25 output. Since then the output price moved from $0.25 to $0.20 per 1M tokens (5 changes).
FAQ
How much does GLM 5.3 Flash cost?
GLM 5.3 Flash costs $0.048 per 1 million input tokens and $0.20 per 1 million output tokens, with cached input at $0.006 (list price on Oct 18, 2026). In practice one chat answer costs about 0.04¢ and summarizing a 20-page PDF about 0.10¢.
Where is GLM 5.3 Flash cheapest?
Among 34 providers we track, the lowest list price today is at OpenInference: $0.048 input and $0.20 output per 1M tokens.
Does GLM 5.3 Flash 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 5.3 Flash cost per task measured or estimated?
Measured: we ran all 13 benchmark prompts on GLM 5.3 Flash through OpenRouter and recorded the tokens it used.