Kimi K3 vs Qwen3.8 Max Prime: cost per task
Kimi K3 is cheaper on 9 of 14 tasks and about 1.1x cheaper on average (geometric mean).
| Task | Kimi K3 | Qwen3.8 Max Prime | Difference |
|---|---|---|---|
| Answer a chat question | 1.6¢ | 2.0¢ | 1.2x |
| Reply to a customer email | 4.2¢ | 0.95¢ | 4.4x |
| Write a 1,500-word blog post | 9.2¢ | 19¢ | 2.1x |
| Write 10 product descriptions | 3.7¢ | 19¢ | 5.3x |
| Translate 1,000 words | 14¢ | 2.1¢ | 6.8x |
| Proofread one page (500 words) | 4.9¢ | 2.5¢ | 2.0x |
| Summarize a 20-page PDF | 3.0¢ | 8.4¢ | 2.8x |
| Extract data from an invoice | 3.2¢ | 2.8¢ | 1.1x |
| Analyze a 200-row spreadsheet | 13¢ | 22¢ | 1.7x |
| Answer from 5 retrieved documents (RAG) | 3.9¢ | 5.0¢ | 1.3x |
| Classify 100 support tickets | 8.8¢ | 11¢ | 1.2x |
| Review 500 lines of code | 11¢ | 21¢ | 1.9x |
| Write a SQL query | 3.8¢ | 3.2¢ | 1.2x |
| Run a 50-step coding agent | $2.24 | $2.57 | 1.1x |
Specs and list prices
| Kimi K3 | Qwen3.8 Max Prime | |
|---|---|---|
| Input per 1M | $0.30 | $4.00 |
| Output per 1M | $11.94 | $12.00 |
| Cached input per 1M | $0.21 | $0.50 |
| Context | 1M | 1M |
| Providers | 26 | 1 |
| Reasoning | default | always |
| LMArena Text | 1487 | - |
| Released | Jul 16, 2026 | Sep 23, 2026 |
FAQ
Is Kimi K3 or Qwen3.8 Max Prime cheaper?
Kimi K3 is cheaper on 9 of 14 tasks and about 1.1x cheaper on average (geometric mean).
How much does Kimi K3 cost compared to Qwen3.8 Max Prime per token?
Kimi K3: $0.30 input and $11.94 output per 1M tokens. Qwen3.8 Max Prime: $4.00 input and $12.00 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.