Datasets:

dispatchAI
/

efficiency-ranking

rank int64 1 31	model stringlengths 18 35	cpu_tps float64 2.6 59.7	size_mb float64 1 4.47k	efficiency_score float64 0 59.7	verdict stringclasses 3 values
1	SmolLM2-135M-Instruct-mobile	59.7	1	59.7	BEST
2	Qwen2.5-0.5B-Instruct-mobile-int4	12.5	1	12.5	BEST
3	Llama-3.2-1B-Instruct-Q4-mobile	11.3	1	11.3	BEST
4	SmolLM2-360M-Instruct-mobile	29.1	258.1	0.113	GOOD
5	MiniCPM-V-4.6-mobile	23.4	504.6	0.046	GOOD
6	Qwen2.5-0.5B-Chinese-mobile	16.4	400.6	0.041	GOOD
7	Qwen2.5-0.5B-Instruct-Server-mobile	16.7	482.3	0.035	GOOD
8	MiniCPM5-1B-multilingual-mobile	18.4	656.2	0.028	GOOD
9	TinyLlama-1.1B-Chat-mobile-int4	17.6	637.8	0.028	GOOD
10	MiniCPM5-1B-mobile	18.1	656.2	0.028	GOOD
11	Qwen2.5-0.5B-Coder-mobile	8	379.4	0.021	OK
12	SmolLM2-1.7B-Instruct-mobile	18.7	1,006.7	0.019	OK
13	Qwen2.5-Math-1.5B-mobile	15.7	940.4	0.017	OK
14	TinyLlama-1.1B-Chat-Q5-mobile	10.9	746.7	0.015	OK
15	Gemma-2B-Arabic-mobile	14.8	1,554.7	0.01	OK
16	SmolLM2-1.7B-Instruct-Q5-mobile	10.8	1,168.7	0.009	OK
17	Qwen2.5-Coder-1.5B-mobile	7.8	940.4	0.008	OK
18	Qwen2.5-1.5B-Instruct-mobile-int4	7.7	1,065.6	0.007	OK
19	Qwen2.5-1.5B-Instruct-Q8-mobile	8.7	1,570.3	0.006	OK
20	Llama-3.2-1B-Instruct-Q6-mobile	5	974.5	0.005	OK
21	Gemma-2-2B-IT-mobile	8.2	1,629.4	0.005	OK
22	Llama-3.2-3B-FunctionCall-mobile	9.1	1,925.8	0.005	OK
23	Llama-3.2-1B-FunctionCall-mobile	8.9	1,925.8	0.005	OK
24	Qwen2.5-1.5B-Instruct-Q5-mobile	4.4	1,072.9	0.004	OK
25	Llama-3.2-1B-Instruct-mobile	4.9	1,259.9	0.004	OK
26	Phi-3.5-mini-Instruct-mobile	8.6	2,282.4	0.004	OK
27	Gemma-2-2B-IT-Q5-mobile	5.7	1,834.2	0.003	OK
28	Phi-3.5-mini-instruct-Q5-mobile	7	2,684.9	0.003	OK
29	Llama-3.2-3B-Instruct-mobile	4.8	1,925.8	0.002	OK
30	Llama-3.2-3B-Instruct-Q5-mobile	2.6	2,214.6	0.001	OK
31	Qwen2.5-Coder-7B-mobile	3	4,466.1	0.001	OK

Efficiency Ranking

Models ranked by tokens-per-second per MB of file size. Higher = more efficient.

🚀 dispatchAI

Downloads last month: -

Number of rows:

Total file size:

7.81 kB