Includes 693 hourly rounds up to 2026-11-16 from NIMStats by Mauro Druwel (source), measured with a different prompt and timeout. Our own probe adds a round every 30 minutes.
Model comparison
Head-to-head from live probes, last 24 hours.
Head-to-head stats
muse-glimmer-30b leads 6 rows, gemma-4-31b-it leads 1.
| muse-glimmer-30b | Metric | gemma-4-31b-it |
|---|---|---|
| meta Serving | Status | Unstable google |
| 65● | Score | 35 |
| 92.3%● | Uptime | 38.5% |
| 26.4 s● | First token (median) | 59.1 s |
| 52.4 s● | Full reply (median) | 66.2 s |
| 19.6 s● | Best reply | 33.2 s |
| 8.2 tok/s | Throughput (median) | ●31.8 tok/s |
| 0 | Round wins | 0 |
| #9● | Leaderboard rank | #11 |
| 60.0% | Faster when both answered (5 rounds) | 40.0% |
marks the better value in each row
Full reply time overlay
Each probe, both models on one axis. Lower is better.
Win timeline
Per round, oldest left: which model replied faster.
BAAA·AAAABAAA
Amuse-glimmer-30b Bgemma-4-31b-it ·neither answered