GLM 5.3 Flash vs GPT-6 Astra: cost per task
GLM 5.3 Flash is cheaper on 14 of 14 tasks and about 87.6x cheaper on average (geometric mean). On LMArena Text they are practically tied (1477 vs 1474, within the margin of error).
| Task | GLM 5.3 Flash | GPT-6 Astra | Difference |
|---|---|---|---|
| Answer a chat question | 0.04¢ | 3.4¢ | 93.7x |
| Reply to a customer email | <0.01¢ | 0.89¢ | 138.1x |
| Write a 1,500-word blog post | 0.14¢ | 11¢ | 82.9x |
| Write 10 product descriptions | 0.14¢ | 6.2¢ | 43.3x |
| Translate 1,000 words | 0.21¢ | 8.0¢ | 38.2x |
| Proofread one page (500 words) | 0.02¢ | 4.0¢ | 207.2x |
| Summarize a 20-page PDF | 0.10¢ | 16¢ | 157.4x |
| Extract data from an invoice | 0.02¢ | 2.6¢ | 116.6x |
| Analyze a 200-row spreadsheet | 0.34¢ | 42¢ | 122.7x |
| Answer from 5 retrieved documents (RAG) | 0.06¢ | 5.9¢ | 103.1x |
| Classify 100 support tickets | 0.16¢ | 3.9¢ | 23.6x |
| Review 500 lines of code | 0.34¢ | 24¢ | 71.5x |
| Write a SQL query | 0.03¢ | 2.4¢ | 74.4x |
| Run a 50-step coding agent | 2.6¢ | $3.72 | 145.2x |
Specs and list prices
| GLM 5.3 Flash | GPT-6 Astra | |
|---|---|---|
| Input per 1M | $0.048 | $10.00 |
| Output per 1M | $0.20 | $50.00 |
| Cached input per 1M | $0.006 | $1.00 |
| Context | 1M | 1.05M |
| Providers | 34 | 7 |
| Reasoning | always | always |
| LMArena Text | 1474 | 1477 |
| Released | Aug 26, 2026 | Sep 4, 2026 |
FAQ
Is GLM 5.3 Flash or GPT-6 Astra cheaper?
GLM 5.3 Flash is cheaper on 14 of 14 tasks and about 87.6x cheaper on average (geometric mean). On LMArena Text they are practically tied (1477 vs 1474, within the margin of error).
How much does GLM 5.3 Flash cost compared to GPT-6 Astra per token?
GLM 5.3 Flash: $0.048 input and $0.20 output per 1M tokens. GPT-6 Astra: $10.00 input and $50.00 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.