User profile

rust_witch

3 runsFirst seen 2026-05-08Avg 75.6

Best PipelineScore

83.3MAINLINE

Total tokens17.1KAcross every task this user has run

Avg latency772msPer task, across all submissions

Tasks run753 submissions x ~25 tasks

Rigs used3Distinct hardware tags

Category signature

Average score per category across all 3 runs.

Code

71.2

Reason

74.9

Write

77.9

Tool Use

77.9

RAG

78.2

Speed

77.8

Rigs this user benchmarked on.

m3-max-128gb1 (33%)

rtx-3090-24gb1 (33%)

ryzen-7950x-cpu-only1 (33%)

Where they spend their tokens.

meta2 (67%)

deepseek1 (33%)

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

#	Model	Provider	Best Score	Tier	Achieved
1	DeepSeek V4	deepseek	83.3	MAINLINE	2026-05-21
2	Llama 4 405B	meta	72.0	FEEDER	2026-05-13

Every run, ordered by score.

#	Model	Hardware	Score	Tier	Code	Reason	Write	Tool Use	RAG	Speed	Tokens	Avg ms	Date
1	DeepSeek V4	m3-max-128gb	83.3	MAINLINE	77.9	85.4	87.1	88.3	84.4	79.2	5.3K	760	2026-05-21
2	Llama 4 405B	rtx-3090-24gb	72.0	FEEDER	65.4	72.3	75.8	74.3	73.2	76.0	5.8K	714	2026-05-13
3	Llama 4 405B	ryzen-7950x-cpu-only	71.6	FEEDER	70.2	67.1	70.9	71.0	76.9	78.1	6.1K	843	2026-05-08