Claude Fable 5.1 vs Kimi K3: cost per task
Kimi K3 is cheaper on 11 of 14 tasks and about 2.1x cheaper on average (geometric mean). Claude Fable 5.1 has the higher LMArena Text score (1502 vs 1487).
| Task | Claude Fable 5.1 | Kimi K3 | Difference |
|---|---|---|---|
| Answer a chat question | 6.3¢ | 1.6¢ | 3.9x |
| Reply to a customer email | 2.0¢ | 4.2¢ | 2.1x |
| Write a 1,500-word blog post | 18¢ | 9.2¢ | 2.0x |
| Write 10 product descriptions | 22¢ | 3.7¢ | 6.0x |
| Translate 1,000 words | 13¢ | 14¢ | 1.1x |
| Proofread one page (500 words) | 5.6¢ | 4.9¢ | 1.1x |
| Summarize a 20-page PDF | 22¢ | 3.0¢ | 7.5x |
| Extract data from an invoice | 5.2¢ | 3.2¢ | 1.6x |
| Analyze a 200-row spreadsheet | 85¢ | 13¢ | 6.7x |
| Answer from 5 retrieved documents (RAG) | 14¢ | 3.9¢ | 3.5x |
| Classify 100 support tickets | 4.5¢ | 8.8¢ | 2.0x |
| Review 500 lines of code | 40¢ | 11¢ | 3.6x |
| Write a SQL query | 5.5¢ | 3.8¢ | 1.5x |
| Run a 50-step coding agent | $3.46 | $2.24 | 1.5x |
Specs and list prices
| Claude Fable 5.1 | Kimi K3 | |
|---|---|---|
| Input per 1M | $10.00 | $0.30 |
| Output per 1M | $50.00 | $11.94 |
| Cached input per 1M | $0.25 | $0.21 |
| Context | 1M | 1M |
| Providers | 4 | 26 |
| Reasoning | always | default |
| LMArena Text | 1502 | 1487 |
| Released | Sep 1, 2026 | Jul 16, 2026 |
FAQ
Is Claude Fable 5.1 or Kimi K3 cheaper?
Kimi K3 is cheaper on 11 of 14 tasks and about 2.1x cheaper on average (geometric mean). Claude Fable 5.1 has the higher LMArena Text score (1502 vs 1487).
How much does Claude Fable 5.1 cost compared to Kimi K3 per token?
Claude Fable 5.1: $10.00 input and $50.00 output per 1M tokens. Kimi K3: $0.30 input and $11.94 output. Token prices alone mislead - models use different numbers of tokens for the same task, which is what the task table shows.