Claude Fable 5.1 commercial tested
Ninth on the extended CymraegBench leaderboard with an overall score of 80.53.
The facts
| Developer | Anthropic |
|---|---|
| OpenRouter ID | anthropic/claude-fable-5.1 |
| Test pricing | $10 / $50 per million tokens (input/output) |
| Context | 1 million tokens; up to 128,000 output tokens |
| Full run cost | $7.31 |
CymraegBench result
| Overall score | Welsh reasoning | Practical Welsh | Translation BLEU |
|---|---|---|---|
| 80.53 | 86.05 | 75.00 | 71.07 |
The same sample was used as for the other leading models: up to 100 cases from each of the 19 core suites, with reasoning_effort=low. The 232 CCC cases do not contribute to the overall score.
Fable 5.1 ranks ninth on the extended leaderboard: 0.20 points behind MiMo V2.6 Pro and 0.72 points ahead of Qwen3.8 Flash. It completed 2,052 of the 2,053 cases.
Note on invalid output: one PIQA case returned an empty answer. It remained empty after retries with limits of 2,048, 8,192 and 16,384 tokens, so scored zero. There were no unresolved infrastructure errors.
CCC Language Audit
| Subset | Score |
|---|---|
| Correction and explanation | 93.91 |
| Idiom | 100.00 |
| Terminology | 93.75 |
| Summarisation and quotation | 92.92 |
| References | 76.56 |
| Welsh prompt | 100.00 |
| English prompt | 100.00 |
| Average across the seven subsets | 93.88 |
Fable 5.1 scored 96.88 for correction and 63.28 for explanation. On the references test it scored 78.13 for selecting references, 65.63 for answer content, and 100 for not inventing a source.
See the full leaderboard, CCC Language Audit, or download the summary, scores by suite, CCC dimensions and the metadata.