Evaluations → Gemini 4 Argon
Gemini 4 Argon commercial tested
Third on the extended CymraegBench leaderboard with an overall score of 87.41.
The facts
| Developer | |
|---|---|
| OpenRouter ID | google/gemini-4-argon |
| Test pricing | $2 / $10 per million tokens (input/output) |
| Context | 1,048,576 tokens; up to 65,536 output tokens |
| Full run cost | $0.95 |
CymraegBench result
| Overall score | Welsh reasoning | Practical Welsh | Translation BLEU |
|---|---|---|---|
| 87.41 | 90.15 | 84.67 | 73.66 |
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 model completed all 2,053 cases without invalid output or unresolved infrastructure errors.
Gemini 4 Argon ranks third: 0.14 points behind Claude Opus 5.5 and 1.18 points ahead of Muse Wonder.
CCC Language Audit
| Subset | Score |
|---|---|
| Correction and explanation | 92.73 |
| Idiom | 100.00 |
| Terminology | 96.88 |
| Summarisation and quotation | 94.94 |
| References | 90.63 |
| Welsh prompt | 100.00 |
| English prompt | 100.00 |
| Average across the seven subsets | 96.45 |
Gemini scored 95.31 for correction and 59.38 for explanation. On the references test it scored 100 for selecting references, 68.75 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.