TinySwallow-1.5B-Instruct: benchmark results
Provider: Other. Access: Open.
Unified ELO 1450 ± 44, rank #1525 of 2656 rated models, from 18 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| pfgen-bench - Completion Mode - Fluency | 0.66 | Fluency Score | 64.4 |
| pfgen-bench - Completion Mode - Truthfulness | 0.81 | Truthfulness Score | 64.2 |
| pfgen-bench - Completion Mode - Score | 0.55 | pfgen Score (mean of three) | 63.3 |
| pfgen-bench - QA Mode - Truthfulness | 0.8 | Truthfulness Score | 60.6 |
| pfgen-bench - Completion Mode - Helpfulness | 0.18 | Helpfulness Score | 60.4 |
| pfgen-bench - QA Mode - Fluency | 0.62 | Fluency Score | 57.6 |
| pfgen-bench - QA Mode - Score | 0.53 | pfgen Score (mean of three) | 54.2 |
| pfgen-bench - QA Mode - Helpfulness | 0.17 | Helpfulness Score | 48.8 |
| Ebisu - JF-TE - F1 | 19.37 | Maximal-Term F1 (%) | 42.9 |
| pfgen-bench - Chat Mode - Fluency | 0.59 | Fluency Score | 39.7 |
| Ebisu - JF-TE - Hit Rate@1 | 2.64 | Hit Rate@1 (%) | 38.1 |
| Ebisu - JF-TE - Hit Rate@5 | 7.12 | Hit Rate@5 (%) | 38.1 |
Interactive version: theaggregate.ai/model?slug=tinyswallow-1-5b-instruct · How It Works · Data refreshed daily, snapshot 2026-09-19.