Leaderboard
On-device LLM performance rankings powered by Glicko-2
P30 Pro
AndroidRank
#171
Rating
1,393
±18 RD
Win Rate
39.6%
Conservative Rating
1,357
TG Rating
1,412
PP Rating
1,245
Matches
840
Record
333W – 507L
Models Tested
| Model | TG Median (tok/s) | PP Median (tok/s) | TG Best | PP Best | Runs |
|---|---|---|---|---|---|
| llama-3.2-1b-instruct-q8_0 | 11.36 | 36.24 | 11.36 | 36.24 | 1 |
| qwen2.5-1.5b-instruct-q8_0 | 8.81 | 22.32 | 9.61 | 29.62 | 2 |
| DeepSeek-R1-Distill-Qwen-1.5B-IQ2_M | 7.74 | 10.78 | 7.74 | 10.78 | 1 |
| qwen2.5-3b-instruct-q5_k_m | 6.03 | 9.93 | 6.03 | 9.93 | 1 |
| Llama-3.2-3B-Instruct-Q6_K | 3.28 | 6.68 | 3.28 | 6.68 | 1 |
| Gemmasutra-Mini-2B-v1-Q6_K | 0.15 | 7.99 | 0.15 | 7.99 | 1 |
Head-to-Head Record
1–50 of 207 rows
1 / 5
Performance by App Version
ImprovedRegressed