BenchmarksSource dated 15 Oct 2026
Timestral-123M on GIFT-Eval
Timestral-123M: GIFT-Eval MASE 0.659, vendor reported.
At a glance
- Model variant
- Timestral-123M
- Benchmark
- GIFT-Eval version not disclosed
- Score
- 0.659 ratio
- Evaluator
- Mistral AI
- Provenance
- vendor
- Sample size
- Unknown
- Observed date
- 25 Oct 2026
- Cohort basis
- Same vendor-reported table; incomplete run configuration prevents verified ranking
- Uncertainty
- {'note': 'No confidence intervals or run date supplied', 'reported': False}
Evaluation configuration
- Comparison Scope
- Only these four vendor-reported observations from the same table; not an independently verified leaderboard.
- Mode
- Pretrained zero-shot; single checkpoint; no ensembling. Scores transcribed from one vendor model-card table.