User profile

quant-or-die

18 runsFirst seen 2026-04-12Avg 66.1

Best PipelineScore

85.3MAINLINE

on DeepSeek Coder V2 236B

Total tokens99.9KAcross every task this user has run

Avg latency843msPer task, across all submissions

Tasks run45018 submissions x ~25 tasks

Rigs used14Distinct hardware tags

Category signature

Average score per category across all 18 runs.

Code

64.9

Reason

67.2

Write

65.2

Tool Use

65.9

RAG

68.7

Speed

65.7

Hardware mix

Rigs this user benchmarked on.

b200-192gb2 (11%)

rtx-4080-16gb2 (11%)

rtx-4070-12gb2 (11%)

m3-air-16gb2 (11%)

a100-80gb1 (6%)

macbook-pro-m3-pro-18gb1 (6%)

h100-80gb1 (6%)

h200-141gb1 (6%)

cloud-api1 (6%)

2x-rtx-40901 (6%)

m3-pro-36gb1 (6%)

m2-pro-16gb1 (6%)

snapdragon-x-elite-32gb1 (6%)

rtx-3060-12gb1 (6%)

Provider mix

Where they spend their tokens.

alibaba6 (33%)

meta3 (17%)

community2 (11%)

tii2 (11%)

deepseek1 (6%)

nous1 (6%)

databricks1 (6%)

google1 (6%)

bigcode1 (6%)

Models tried

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

#	Model	Provider	Best Score	Tier	Achieved
1	DeepSeek Coder V2 236B	deepseek	85.3	MAINLINE	2026-04-13
2	Hermes 3 Llama 3.1 405B	nous	84.9	MAINLINE	2026-05-04
3	Llama 4 70B Instruct	meta	81.9	MAINLINE	2026-05-15
4	Qwen 3 14B Instruct	alibaba	79.9	MAINLINE	2026-05-12
5	Qwen 2.5 VL 72B	alibaba	76.9	MAINLINE	2026-05-18
6	Llama 3.1 70B Instruct	meta	76.0	MAINLINE	2026-05-09
7	DBRX Instruct 132B-MoE	databricks	73.8	FEEDER	2026-04-12
8	Qwen 2.5 Coder 7B	alibaba	70.6	FEEDER	2026-04-15
9	L3 70B Euryale	community	69.5	FEEDER	2026-04-29
10	CodeGemma 7B	google	66.2	FEEDER	2026-05-23
11	Falcon Mamba 7B	tii	64.2	FEEDER	2026-05-23
12	StarCoder2 7B	bigcode	62.1	FEEDER	2026-05-18
13	Qwen 2.5 1.5B Instruct	alibaba	43.8	TAP	2026-05-11
14	Llama 3.2 1B Instruct	meta	41.4	TAP	2026-05-11
15	TinyLlama 1.1B Chat	community	29.6	DRIP	2026-05-14

All submissions

Every run, ordered by score.

#	Model	Hardware	Score	Tier	Code	Reason	Write	Tool Use	RAG	Speed	Tokens	Avg ms	Date
1	DeepSeek Coder V2 236B	a100-80gb	85.3	MAINLINE	83.0	85.0	84.1	83.9	89.1	89.8	5.7K	831	2026-04-13
2	Hermes 3 Llama 3.1 405B	b200-192gb	84.9	MAINLINE	85.3	81.5	83.7	87.3	84.0	88.9	5.6K	877	2026-05-04
3	Llama 4 70B Instruct	b200-192gb	81.9	MAINLINE	84.7	81.7	80.5	76.3	87.0	80.4	6.1K	891	2026-05-15
4	Qwen 3 14B Instruct	rtx-4080-16gb	79.9	MAINLINE	80.4	79.1	75.5	77.8	82.9	84.6	5.7K	758	2026-05-12
5	Qwen 3 14B Instruct	macbook-pro-m3-pro-18gb	79.0	MAINLINE	79.1	79.8	78.0	72.2	83.9	82.0	5.6K	765	2026-05-01
6	Qwen 2.5 VL 72B	h100-80gb	76.9	MAINLINE	74.8	78.5	74.7	78.4	78.5	77.7	5.7K	837	2026-05-18
7	Llama 3.1 70B Instruct	h200-141gb	76.0	MAINLINE	72.4	78.1	79.0	75.4	76.1	76.8	4.8K	944	2026-05-09
8	DBRX Instruct 132B-MoE	cloud-api	73.8	FEEDER	72.3	76.4	76.2	75.8	74.8	67.1	5.5K	871	2026-04-12
9	Qwen 2.5 Coder 7B	rtx-4070-12gb	70.6	FEEDER	64.9	74.7	68.4	76.4	69.6	71.8	5.2K	811	2026-04-15
10	L3 70B Euryale	2x-rtx-4090	69.5	FEEDER	68.7	74.5	72.9	63.2	73.7	62.7	5.9K	3016	2026-04-29
11	CodeGemma 7B	rtx-4080-16gb	66.2	FEEDER	66.8	63.7	62.6	68.0	70.5	67.3	5.4K	808	2026-05-23
12	Falcon Mamba 7B	rtx-4070-12gb	64.2	FEEDER	62.0	62.1	60.7	68.7	71.8	63.1	5.4K	787	2026-05-23
13	StarCoder2 7B	m3-pro-36gb	62.1	FEEDER	61.0	63.4	64.0	59.3	67.9	58.0	6.3K	818	2026-05-18
14	Falcon Mamba 7B	m2-pro-16gb	61.6	FEEDER	62.5	64.2	58.6	58.5	64.5	60.4	4.9K	819	2026-05-20
15	Qwen 2.5 1.5B Instruct	snapdragon-x-elite-32gb	43.8	TAP	38.0	43.8	38.3	49.5	49.6	49.2	6.0K	355	2026-05-11
16	Qwen 2.5 1.5B Instruct	m3-air-16gb	43.2	TAP	42.1	44.3	40.3	49.4	42.6	40.5	5.7K	351	2026-05-21
17	Llama 3.2 1B Instruct	m3-air-16gb	41.4	TAP	38.8	46.4	43.0	42.8	37.0	39.4	5.3K	324	2026-05-11
18	TinyLlama 1.1B Chat	rtx-3060-12gb	29.6	DRIP	31.2	32.1	33.0	23.8	32.5	23.1	5.3K	312	2026-05-14