PipelineScore
← Back to user leaderboard
User profile

cuda-cult

11 runsFirst seen 2026-04-15Avg 67.9
Best PipelineScore
77.4MAINLINE
Total tokens61.0KAcross every task this user has run
Avg latency1319msPer task, across all submissions
Tasks run27511 submissions x ~25 tasks
Rigs used7Distinct hardware tags

Category signature

Average score per category across all 11 runs.

Code
68.1
Reason
68.8
Write
68.0
Tool Use
67.9
RAG
68.3
Speed
65.5

Hardware mix

Rigs this user benchmarked on.

ryzen-7950x-cpu-only4 (36%)
m3-ultra-256gb2 (18%)
2x-rtx-40901 (9%)
rtx-3060-12gb1 (9%)
m3-pro-36gb1 (9%)
macbook-pro-m3-pro-18gb1 (9%)
m1-air-8gb1 (9%)

Provider mix

Where they spend their tokens.

alibaba3 (27%)
cohere2 (18%)
community1 (9%)
meta1 (9%)
google1 (9%)
internlm1 (9%)
cognitivecomputations1 (9%)
microsoft1 (9%)

Models tried

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

#ModelBest ScoreTier
1Command A77.4MAINLINE
2Qwen 2.5 7B Instruct70.8FEEDER
3L3 70B Euryale70.5FEEDER
4Qwen 2.5 Coder 7B68.7FEEDER
5Llama 4 8B Instruct68.2FEEDER
6Gemma 2 9B IT66.8FEEDER
7InternLM 2.5 7B Chat65.8FEEDER
8Dolphin 2.9 Llama 3 8B62.9FEEDER
9Qwen 3 4B Instruct61.1FEEDER
10Phi 3.5 Mini58.0TAP

All submissions

Every run, ordered by score.

#ModelScoreTier
1Command A77.4MAINLINE
2Command A76.5MAINLINE
3Qwen 2.5 7B Instruct70.8FEEDER
4L3 70B Euryale70.5FEEDER
5Qwen 2.5 Coder 7B68.7FEEDER
6Llama 4 8B Instruct68.2FEEDER
7Gemma 2 9B IT66.8FEEDER
8InternLM 2.5 7B Chat65.8FEEDER
9Dolphin 2.9 Llama 3 8B62.9FEEDER
10Qwen 3 4B Instruct61.1FEEDER
11Phi 3.5 Mini58.0TAP
cuda-cult · PipelineScore