PipelineScore
โ† Back to user leaderboard
User profile

lab

96 runsFirst seen 2026-04-10Avg 68.0
Best PipelineScore
89.5MAINLINE
Total tokens528.9KAcross every task this user has run
Avg latency1000msPer task, across all submissions
Tasks run2.4K96 submissions x ~25 tasks
Rigs used28Distinct hardware tags

Category signature

Average score per category across all 96 runs.

Code
68.2
Reason
67.8
Write
68.3
Tool Use
68.5
RAG
68.2
Speed
66.5

Hardware mix

Rigs this user benchmarked on.

lab-macbook-pro-m3-pro-18gb13 (14%)
lab-ryzen-7950x-cpu-only9 (9%)
lab-dgx-h1006 (6%)
lab-rtx-3060-12gb6 (6%)
lab-baseline5 (5%)
lab-rtx-3080-10gb5 (5%)
lab-h200-141gb4 (4%)
lab-m2-ultra-192gb4 (4%)
lab-a100-40gb4 (4%)
lab-m2-pro-16gb4 (4%)
lab-m3-max-64gb3 (3%)
lab-rtx-3090-24gb3 (3%)
lab-m3-pro-36gb3 (3%)
lab-m3-air-16gb3 (3%)
lab-m2-air-16gb3 (3%)
lab-b200-192gb2 (2%)
lab-m3-ultra-256gb2 (2%)
lab-h100-80gb2 (2%)
lab-4x-rtx-30902 (2%)
lab-a6000-48gb2 (2%)
lab-m3-max-128gb2 (2%)
lab-rtx-4080-16gb2 (2%)
lab-rtx-4070-12gb2 (2%)
lab-rtx-4080-16gb-offload1 (1%)
lab-a100-80gb1 (1%)
lab-ryzen-7950x-rtx-30901 (1%)
lab-snapdragon-x-elite-32gb1 (1%)
lab-m1-air-8gb1 (1%)

Provider mix

Where they spend their tokens.

alibaba19 (20%)
meta9 (9%)
deepseek8 (8%)
mistral8 (8%)
google7 (7%)
microsoft6 (6%)
cohere5 (5%)
community5 (5%)
yi4 (4%)
nous3 (3%)
bigcode3 (3%)
zhipu2 (2%)
internlm2 (2%)
cognitivecomputations2 (2%)
tii2 (2%)
ibm2 (2%)
huggingface2 (2%)
moonshot1 (1%)
databricks1 (1%)
xai1 (1%)
openchat1 (1%)
openbmb1 (1%)
upstage1 (1%)
zyphra1 (1%)

Models tried

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

#ModelBest ScoreTier
1DeepSeek R1 671B-A37B89.5MAINLINE
2DeepSeek Coder V2 236B87.3MAINLINE
3Qwen 3 72B Instruct85.0MAINLINE
4Qwen 3 235B-A22B MoE84.0MAINLINE
5DeepSeek V483.7MAINLINE
6Hermes 3 Llama 3.1 405B83.1MAINLINE
7Qwen 3 32B Instruct82.4MAINLINE
8Command A82.0MAINLINE
9Llama 4 70B Instruct81.2MAINLINE
10Mixtral 8x22B Instruct81.2MAINLINE
11DeepSeek R1 Distill Qwen 32B81.2MAINLINE
12DeepSeek V3 671B-A37B80.8MAINLINE
13WizardLM 2 8x22B80.0MAINLINE
14Qwen 2.5 72B Instruct79.9MAINLINE
15Kimi K2 Instruct79.8MAINLINE
16Qwen 3.6 72B79.4MAINLINE
17Qwen 2.5 VL 72B78.8MAINLINE
18Llama 3.3 70B Instruct78.7MAINLINE
19DeepSeek V2.578.6MAINLINE
20Mistral Large 278.5MAINLINE
21Qwen 2.5 Coder 32B78.5MAINLINE
22Hermes 3 Llama 3.1 70B78.0MAINLINE
23GLM 4 Plus77.9MAINLINE
24Codestral 22B77.9MAINLINE
25Qwen 2.5 14B Instruct76.9MAINLINE
26Llama 3.1 70B Instruct76.8MAINLINE
27Devstral Small 24B76.7MAINLINE
28Qwen 2.5 32B Instruct76.3MAINLINE
29Yi 1.5 34B Chat75.8MAINLINE
30DeepSeek R1 Distill Llama 8B75.3MAINLINE
31Qwen 3 14B Instruct74.2FEEDER
32Llama 4 405B73.6FEEDER
33DeepSeek Coder V2 16B73.4FEEDER
34Gemma 3 27B IT73.4FEEDER
35Gemma 2 27B IT73.3FEEDER
36DBRX Instruct 132B-MoE73.1FEEDER
37StarCoder2 15B72.7FEEDER
38Qwen 3 8B Instruct72.6FEEDER
39Phi 4 14B72.3FEEDER
40Gemma 3 12B IT72.2FEEDER
41Command R+71.8FEEDER
42Magnum V4 72B71.6FEEDER
43L3 70B Euryale71.0FEEDER
44Qwen 2.5 Coder 7B70.7FEEDER
45Mistral Small 24B Instruct70.7FEEDER
46Aya 23 35B70.0FEEDER
47Hermes 3 Llama 3.1 8B70.0FEEDER
48Mistral Nemo 12B Instruct69.4FEEDER
49GLM 4 9B Chat69.0FEEDER
50Mixtral 8x7B Instruct68.7FEEDER
51Phi 3.5 MoE 42B68.4FEEDER
52Yi Coder 9B68.4FEEDER
53Llama 4 8B Instruct68.3FEEDER
54InternLM 2.5 20B Chat68.3FEEDER
55Grok-1 314B68.0FEEDER
56Command R67.8FEEDER
57OpenChat 3.6 8B67.7FEEDER
58InternLM 2.5 7B Chat66.9FEEDER
59Gemma 2 9B IT66.8FEEDER
60MiniCPM-V 2.6 8B66.7FEEDER
61LLaVA OneVision 7B66.3FEEDER
62SOLAR 10.7B Instruct65.6FEEDER
63Qwen 2.5 VL 7B65.3FEEDER
64Yi 1.5 9B Chat65.1FEEDER
65Dolphin 3.0 Llama 3.1 8B65.1FEEDER
66Dolphin 2.9 Llama 3 8B65.0FEEDER
67Falcon 3 10B Instruct64.8FEEDER
68Qwen 3 4B Instruct64.6FEEDER
69Qwen 2.5 7B Instruct64.4FEEDER
70Granite 3.1 8B Instruct64.3FEEDER
71Code Llama 34B Instruct64.0FEEDER
72Llama 3.1 8B Instruct64.0FEEDER
73Phi 3 Small 7B62.9FEEDER
74StarCoder2 7B62.4FEEDER
75CodeGemma 7B62.2FEEDER
76Falcon Mamba 7B61.9FEEDER
77Zamba 2 7B Instruct60.4FEEDER
78Gemma 3 4B IT59.8TAP
79Aya 23 8B59.3TAP
80Yi 1.5 6B Chat57.2TAP
81Mistral 7B Instruct v0.357.1TAP
82LLaVA 1.6 Mistral 7B56.7TAP
83Phi 3.5 Mini56.0TAP
84Qwen 2.5 3B Instruct55.1TAP
85Phi 3 Mini 3.8B55.1TAP
86Qwen 2.5 Coder 1.5B53.8TAP
87Llama 3.2 3B Instruct50.1TAP
88Gemma 2 2B IT49.5TAP
89Granite 3.1 2B Instruct49.3TAP
90StarCoder2 3B46.1TAP
91Qwen 2.5 1.5B Instruct43.4TAP
92SmolLM2 1.7B40.6TAP
93Llama 3.2 1B Instruct36.5DRIP
94SmolLM 1.7B32.7DRIP
95Qwen 2.5 0.5B Instruct32.5DRIP
96TinyLlama 1.1B Chat28.9DRIP

All submissions

Every run, ordered by score.

#ModelScoreTier
1DeepSeek R1 671B-A37B89.5MAINLINE
2DeepSeek Coder V2 236B87.3MAINLINE
3Qwen 3 72B Instruct85.0MAINLINE
4Qwen 3 235B-A22B MoE84.0MAINLINE
5DeepSeek V483.7MAINLINE
6Hermes 3 Llama 3.1 405B83.1MAINLINE
7Qwen 3 32B Instruct82.4MAINLINE
8Command A82.0MAINLINE
9Llama 4 70B Instruct81.2MAINLINE
10Mixtral 8x22B Instruct81.2MAINLINE
11DeepSeek R1 Distill Qwen 32B81.2MAINLINE
12DeepSeek V3 671B-A37B80.8MAINLINE
13WizardLM 2 8x22B80.0MAINLINE
14Qwen 2.5 72B Instruct79.9MAINLINE
15Kimi K2 Instruct79.8MAINLINE
16Qwen 3.6 72B79.4MAINLINE
17Qwen 2.5 VL 72B78.8MAINLINE
18Llama 3.3 70B Instruct78.7MAINLINE
19DeepSeek V2.578.6MAINLINE
20Mistral Large 278.5MAINLINE
21Qwen 2.5 Coder 32B78.5MAINLINE
22Hermes 3 Llama 3.1 70B78.0MAINLINE
23GLM 4 Plus77.9MAINLINE
24Codestral 22B77.9MAINLINE
25Qwen 2.5 14B Instruct76.9MAINLINE
26Llama 3.1 70B Instruct76.8MAINLINE
27Devstral Small 24B76.7MAINLINE
28Qwen 2.5 32B Instruct76.3MAINLINE
29Yi 1.5 34B Chat75.8MAINLINE
30DeepSeek R1 Distill Llama 8B75.3MAINLINE
31Qwen 3 14B Instruct74.2FEEDER
32Llama 4 405B73.6FEEDER
33DeepSeek Coder V2 16B73.4FEEDER
34Gemma 3 27B IT73.4FEEDER
35Gemma 2 27B IT73.3FEEDER
36DBRX Instruct 132B-MoE73.1FEEDER
37StarCoder2 15B72.7FEEDER
38Qwen 3 8B Instruct72.6FEEDER
39Phi 4 14B72.3FEEDER
40Gemma 3 12B IT72.2FEEDER
41Command R+71.8FEEDER
42Magnum V4 72B71.6FEEDER
43L3 70B Euryale71.0FEEDER
44Qwen 2.5 Coder 7B70.7FEEDER
45Mistral Small 24B Instruct70.7FEEDER
46Aya 23 35B70.0FEEDER
47Hermes 3 Llama 3.1 8B70.0FEEDER
48Mistral Nemo 12B Instruct69.4FEEDER
49GLM 4 9B Chat69.0FEEDER
50Mixtral 8x7B Instruct68.7FEEDER
51Phi 3.5 MoE 42B68.4FEEDER
52Yi Coder 9B68.4FEEDER
53Llama 4 8B Instruct68.3FEEDER
54InternLM 2.5 20B Chat68.3FEEDER
55Grok-1 314B68.0FEEDER
56Command R67.8FEEDER
57OpenChat 3.6 8B67.7FEEDER
58InternLM 2.5 7B Chat66.9FEEDER
59Gemma 2 9B IT66.8FEEDER
60MiniCPM-V 2.6 8B66.7FEEDER
61LLaVA OneVision 7B66.3FEEDER
62SOLAR 10.7B Instruct65.6FEEDER
63Qwen 2.5 VL 7B65.3FEEDER
64Yi 1.5 9B Chat65.1FEEDER
65Dolphin 3.0 Llama 3.1 8B65.1FEEDER
66Dolphin 2.9 Llama 3 8B65.0FEEDER
67Falcon 3 10B Instruct64.8FEEDER
68Qwen 3 4B Instruct64.6FEEDER
69Qwen 2.5 7B Instruct64.4FEEDER
70Granite 3.1 8B Instruct64.3FEEDER
71Code Llama 34B Instruct64.0FEEDER
72Llama 3.1 8B Instruct64.0FEEDER
73Phi 3 Small 7B62.9FEEDER
74StarCoder2 7B62.4FEEDER
75CodeGemma 7B62.2FEEDER
76Falcon Mamba 7B61.9FEEDER
77Zamba 2 7B Instruct60.4FEEDER
78Gemma 3 4B IT59.8TAP
79Aya 23 8B59.3TAP
80Yi 1.5 6B Chat57.2TAP
81Mistral 7B Instruct v0.357.1TAP
82LLaVA 1.6 Mistral 7B56.7TAP
83Phi 3.5 Mini56.0TAP
84Qwen 2.5 3B Instruct55.1TAP
85Phi 3 Mini 3.8B55.1TAP
86Qwen 2.5 Coder 1.5B53.8TAP
87Llama 3.2 3B Instruct50.1TAP
88Gemma 2 2B IT49.5TAP
89Granite 3.1 2B Instruct49.3TAP
90StarCoder2 3B46.1TAP
91Qwen 2.5 1.5B Instruct43.4TAP
92SmolLM2 1.7B40.6TAP
93Llama 3.2 1B Instruct36.5DRIP
94SmolLM 1.7B32.7DRIP
95Qwen 2.5 0.5B Instruct32.5DRIP
96TinyLlama 1.1B Chat28.9DRIP
lab ยท PipelineScore