Kimi K2 Thinking is 19.4× cheaper per million tokens (blended). Claude Fable 5 has the larger context window (1,000,000 vs 262,144 tokens, 3.8×). Kimi K2 Thinking is open-weights; the other is not.
Claude Fable 5 vs Kimi K2 Thinking
Live catalog fields. Best-in-row is highlighted.
Claude Fable 5
Anthropic · anthropic__claude-fable-5
Kimi K2 Thinkingopen
Moonshotai · moonshotai__kimi-k2-thinking
Identity
| Field | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
|---|---|---|
| Provider | Anthropic | Moonshotai |
| Slug | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
| Status | live | live |
| Open weights | No | Yes |
| License | — | — |
| HuggingFace | — | moonshotai/Kimi-K2-Thinking |
Quality
| Field | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
|---|---|---|
| Quality index (0–100) | 91.1 | — |
Cost
| Field | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
|---|---|---|
| Input $ / 1M tokens | $10.00 | $0.60 |
| Cached input $ / 1M | $1.00 | $0.15 |
| Output $ / 1M tokens | $50.00 | $2.50 |
Context
| Field | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
|---|---|---|
| Max context tokens | 1,000,000 | 262,144 |
| Max output tokens | 128,000 | 100,352 |
Modalities
| Field | anthropic__claude-fable-5 | moonshotai__kimi-k2-thinking |
|---|---|---|
| Input | text, image, file | text |
| Output | text | text |