Leaderboard
On-device LLM performance rankings powered by Glicko-2
Xiaomi 11T Pro
AndroidRank
#114
Rating
1,557
±14 RD
Win Rate
55.5%
Conservative Rating
1,528
TG Rating
1,570
PP Rating
1,386
Matches
1,300
Record
721W – 579L
Models Tested
| Model | TG Median (tok/s) | PP Median (tok/s) | TG Best | PP Best | Runs |
|---|---|---|---|---|---|
| llama-3.2-1b-instruct-q8_0 | 13.01 | 40.77 | 13.01 | 40.77 | 1 |
| qwen2.5-1.5b-instruct-q8_0 | 9.97 | 32.54 | 10.02 | 33.78 | 2 |
| DeepSeek-R1-Distill-Qwen-1.5B-Q5_K_M | 9.75 | 17.48 | 9.75 | 17.48 | 1 |
| SmolLM2-1.7B-Instruct-Q8_0 | 8.77 | 23.96 | 8.77 | 23.96 | 1 |
| Phi-3.5-mini-instruct.Q4_K_M | 6.22 | 9.93 | 6.22 | 9.93 | 1 |
| qwen2.5-3b-instruct-q5_k_m | 6.00 | 11.32 | 6.00 | 11.32 | 1 |
| Gemmasutra-Mini-2B-v1-Q6_K | 5.73 | 12.71 | 6.15 | 14.25 | 2 |
| gemma-2-2b-it-Q6_K | 5.56 | 12.93 | 7.04 | 14.94 | 2 |
| Llama-3.2-3B-Instruct-Q6_K | 5.35 | 9.10 | 5.35 | 9.10 | 1 |
Head-to-Head Record
1–50 of 315 rows
1 / 7
Performance by App Version
ImprovedRegressed