WizardLM-2 8x22B is 48.4× cheaper per million tokens (blended). Claude Fable 5 has the larger context window (1,000,000 vs 65,535 tokens, 15.3×). WizardLM-2 8x22B is open-weights; the other is not.
Claude Fable 5 vs WizardLM-2 8x22B
Live catalog fields. Best-in-row is highlighted.
Claude Fable 5
Anthropic · anthropic__claude-fable-5
WizardLM-2 8x22Bopen
Microsoft · microsoft__wizardlm-2-8x22b
Identity
| Field | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
|---|---|---|
| Provider | Anthropic | Microsoft |
| Slug | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
| Status | live | live |
| Open weights | No | Yes |
| License | — | — |
| HuggingFace | — | microsoft/WizardLM-2-8x22B |
Quality
| Field | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
|---|---|---|
| Quality index (0–100) | 89.4 | — |
Cost
| Field | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
|---|---|---|
| Input $ / 1M tokens | $10.00 | $0.62 |
| Cached input $ / 1M | $1.00 | — |
| Output $ / 1M tokens | $50.00 | $0.62 |
Context
| Field | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
|---|---|---|
| Max context tokens | 1,000,000 | 65,535 |
| Max output tokens | 128,000 | 8,000 |
Modalities
| Field | anthropic__claude-fable-5 | microsoft__wizardlm-2-8x22b |
|---|---|---|
| Input | text, image, file | text |
| Output | text | text |