Gemma 4 31B is 139.5× cheaper per million tokens (blended). Claude Fable 5 leads quality index by 13.0 points (89.4 vs 76.4). Claude Fable 5 has the larger context window (1,000,000 vs 262,144 tokens, 3.8×). Gemma 4 31B is open-weights; the other is not.
Claude Fable 5 vs Gemma 4 31B
Live catalog fields. Best-in-row is highlighted.
Claude Fable 5
Anthropic · anthropic__claude-fable-5
Gemma 4 31Bopen
Google · google__gemma-4-31b-it
Identity
| Field | anthropic__claude-fable-5 | google__gemma-4-31b-it |
|---|---|---|
| Provider | Anthropic | |
| Slug | anthropic__claude-fable-5 | google__gemma-4-31b-it |
| Status | live | live |
| Open weights | No | Yes |
| License | — | — |
| HuggingFace | — | google/gemma-4-31B-it |
Quality
| Field | anthropic__claude-fable-5 | google__gemma-4-31b-it |
|---|---|---|
| Quality index (0–100) | 89.4 | 76.4 |
Cost
| Field | anthropic__claude-fable-5 | google__gemma-4-31b-it |
|---|---|---|
| Input $ / 1M tokens | $10.00 | $0.09 |
| Cached input $ / 1M | $1.00 | $0.05 |
| Output $ / 1M tokens | $50.00 | $0.34 |
Context
| Field | anthropic__claude-fable-5 | google__gemma-4-31b-it |
|---|---|---|
| Max context tokens | 1,000,000 | 262,144 |
| Max output tokens | 128,000 | 16,384 |
Modalities
| Field | anthropic__claude-fable-5 | google__gemma-4-31b-it |
|---|---|---|
| Input | text, image, file | image, text, video |
| Output | text | text |