User profile

local-llama-fan

15 runsFirst seen 2026-04-10Avg 64.5

Best PipelineScore

84.8MAINLINE

on Hermes 3 Llama 3.1 405B

Total tokens82.7KAcross every task this user has run

Avg latency718msPer task, across all submissions

Tasks run37515 submissions x ~25 tasks

Rigs used9Distinct hardware tags

Category signature

Average score per category across all 15 runs.

Code

64.1

Reason

65.3

Write

63.7

Tool Use

64.6

RAG

65.4

Speed

63.9

Hardware mix

Rigs this user benchmarked on.

rtx-3060-12gb4 (27%)

m2-pro-16gb2 (13%)

macbook-pro-m3-pro-18gb2 (13%)

m3-air-16gb2 (13%)

h100-80gb1 (7%)

h200-141gb1 (7%)

m3-max-128gb1 (7%)

dgx-h1001 (7%)

m3-pro-36gb1 (7%)

Provider mix

Where they spend their tokens.

alibaba4 (27%)

google2 (13%)

mistral2 (13%)

nous1 (7%)

deepseek1 (7%)

cognitivecomputations1 (7%)

yi1 (7%)

zyphra1 (7%)

microsoft1 (7%)

community1 (7%)

Models tried

Best score per model. Click a model to see its full page.

#	Model	Provider	Best Score	Tier	Achieved
1	Hermes 3 Llama 3.1 405B	nous	84.8	MAINLINE	2026-04-26
2	DeepSeek R1 671B-A37B	deepseek	83.0	MAINLINE	2026-04-27
3	Qwen 2.5 32B Instruct	alibaba	76.0	MAINLINE	2026-04-10
4	Gemma 3 12B IT	google	75.7	MAINLINE	2026-04-10
5	Qwen 2.5 14B Instruct	alibaba	75.2	MAINLINE	2026-04-21
6	Mixtral 8x7B Instruct	mistral	69.2	FEEDER	2026-04-29
7	Mistral Nemo 12B Instruct	mistral	68.7	FEEDER	2026-04-17
8	Dolphin 2.9 Llama 3 8B	cognitivecomputations	66.8	FEEDER	2026-05-16
9	Yi 1.5 6B Chat	yi	61.1	FEEDER	2026-04-21
10	Zamba 2 7B Instruct	zyphra	59.5	TAP	2026-05-20
11	Qwen 2.5 3B Instruct	alibaba	58.8	TAP	2026-05-05
12	Qwen 2.5 Coder 1.5B	alibaba	53.3	TAP	2026-05-08
13	Phi 3 Mini 3.8B	microsoft	52.9	TAP	2026-04-23
14	Gemma 2 2B IT	google	51.2	TAP	2026-05-23
15	TinyLlama 1.1B Chat	community	30.8	DRIP	2026-05-22

All submissions

Every run, ordered by score.

#	Model	Hardware	Score	Tier	Code	Reason	Write	Tool Use	RAG	Speed	Tokens	Avg ms	Date
1	Hermes 3 Llama 3.1 405B	h100-80gb	84.8	MAINLINE	83.1	86.7	89.7	81.1	87.2	81.8	5.3K	915	2026-04-26
2	DeepSeek R1 671B-A37B	h200-141gb	83.0	MAINLINE	86.0	79.0	83.0	82.8	82.2	84.7	4.9K	897	2026-04-27
3	Qwen 2.5 32B Instruct	m3-max-128gb	76.0	MAINLINE	71.8	80.4	80.4	73.5	81.8	69.5	4.9K	1563	2026-04-10
4	Gemma 3 12B IT	m2-pro-16gb	75.7	MAINLINE	74.4	75.7	79.2	73.4	72.3	79.6	5.1K	783	2026-04-10
5	Qwen 2.5 14B Instruct	rtx-3060-12gb	75.2	MAINLINE	80.6	73.6	70.7	76.3	75.8	70.6	6.1K	761	2026-04-21
6	Mixtral 8x7B Instruct	dgx-h100	69.2	FEEDER	71.1	65.2	68.2	71.0	71.0	69.4	5.6K	899	2026-04-29
7	Mistral Nemo 12B Instruct	macbook-pro-m3-pro-18gb	68.7	FEEDER	69.8	72.0	64.3	70.7	66.0	67.1	5.7K	808	2026-04-17
8	Dolphin 2.9 Llama 3 8B	m2-pro-16gb	66.8	FEEDER	63.0	69.2	62.4	71.4	68.5	68.9	6.1K	803	2026-05-16
9	Yi 1.5 6B Chat	rtx-3060-12gb	61.1	FEEDER	64.7	58.1	60.7	59.1	60.5	62.5	5.6K	801	2026-04-21
10	Zamba 2 7B Instruct	m3-pro-36gb	59.5	TAP	57.5	58.1	58.9	67.3	57.3	59.6	5.8K	833	2026-05-20
11	Qwen 2.5 3B Instruct	macbook-pro-m3-pro-18gb	58.8	TAP	59.5	60.8	56.1	59.1	62.6	53.6	5.2K	326	2026-05-05
12	Qwen 2.5 Coder 1.5B	rtx-3060-12gb	53.3	TAP	50.9	54.2	56.4	50.6	54.6	54.7	5.6K	349	2026-05-08
13	Phi 3 Mini 3.8B	m3-air-16gb	52.9	TAP	49.4	57.8	52.2	54.4	55.0	49.4	5.6K	354	2026-04-23
14	Gemma 2 2B IT	rtx-3060-12gb	51.2	TAP	50.7	54.8	48.2	45.8	54.6	53.4	5.4K	334	2026-05-23
15	TinyLlama 1.1B Chat	m3-air-16gb	30.8	DRIP	28.5	33.9	25.1	32.0	31.8	34.4	5.6K	339	2026-05-22