Note: Overall leaderboard rankings may not reflect true model quality — individual benchmarks give a clearer picture. ARC-Challenge MMLU GPQA GSM8K Artificial Analysis Intelligence Index v4.1
← Back to leaderboard

diffuspace: clover-35b-a3b-preview

14 benchmarks
Companydiffuspace

Leaderboard position

#274
Overall rank (of 1121 models)
1627
Elo rating
14
Benchmarks
Top 24.4%
Percentile
#1
In diffuspace (of 1 models)
Above: alibaba-ath: happyhorse-1.0 #273 Below: gritlm: gritlm-7b #275

Benchmark highlights

Best benchmark
MATH-500 96.2
Worst benchmark
Terminal-Bench Hard 38.1

Nearest competitor from another company

The model from another company with the most similar benchmark scores.

appeared as nearest in 2 of 14 benchmarks

Meta: muse-glimmer-30b
2/14 14.3%
Model Votes % of benchmarks
Meta: muse-glimmer-30b 2 14.3%
DeepSeek: deepseek-v3.2 1 7.1%
Qwen: Qwen3 8B 1 7.1%
OpenBMB: minicpm5-2b 1 7.1%
Google: Gemma 4 12B Instruct 1 7.1%
Anthropic: claude-3.5-sonnet 1 7.1%
Z.ai: GLM 5 1 7.1%
Xiaomi: MiMo-V2.5 Pro 1 7.1%
Nex AGI: Nex-N2 Mini 1 7.1%
Qwen: Qwen3.5 27B 1 7.1%

Benchmarks

MATH-500 96.2 ZebraLogic 95.8 MMLU-Pro 84.8 GPQA Diamond 83.8 AIME 2026 83.3 SWE-bench Verified 73.6 LiveCodeBench v6 72.6 IFBench 72.3 AA-LCR 72 BFCL v4 67.9 TauBench V3 (Average) 65.3 Terminal-Bench v2.1 51.7 SWE-bench Pro 46.5 Terminal-Bench Hard 38.1

Predicted scores

Estimated from 25 similar models and the global Elo rating. These are projections, not measured results.

Benchmark Predicted Confidence Based on
Humanity's Last Exam (no tools) ~25.03 65% NVIDIA Corporation: nvidia-nemotron-3.5-bolt-74b-a9b-24gb, Qwen: Qwen3.5 27B, Google: Gemma 4 12B Instruct
Artificial Analysis Intelligence Index v4.1 (Maximum Reasoning) ~29.97 65% Z.ai: GLM 5, Qwen: Qwen3.5 27B, Google: Gemma 4 12B Instruct
Artificial Analysis Intelligence Index v4.0 (Maximum Reasoning) ~36.28 65% Z.ai: GLM 5, Qwen: Qwen3.5 27B, LG AI Research: k-exaone-236b-a23b
HMMT Feb 2026 ~80.31 64% Z.ai: GLM 5, Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B
IFEval ~92.61 64% Nex AGI: Nex-N2 Mini, Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B
SciCode ~39.65 64% NVIDIA Corporation: nvidia-nemotron-3.5-bolt-74b-a9b-24gb, Meta: muse-glimmer-30b, ifm: k2-horizon-mova-36b-a4b
MMMU-Pro ~69.15 64% Qwen: Qwen3.5 27B, Google: Gemma 4 12B Instruct, Meta: muse-glimmer-30b
BrowseComp ~62.63 64% Nex AGI: Nex-N2 Mini, NVIDIA Corporation: nvidia-nemotron-3.5-bolt-74b-a9b-24gb, Qwen: Qwen3.5 27B
Terminal-Bench v2.0 ~44.97 64% Z.ai: GLM 5, Qwen: Qwen3.5 27B, Cohere Labs: north-mini-code-1.0
Text Arena (Snapshot from 20 june 2026) ~1414.97 64% Qwen: Qwen3.5 27B, Xiaomi: mimo-v2-pro, Meta: Muse Spark
Claw-Eval ~58.46 63% Qwen: Qwen3.5 27B, xingchen-agi: xing4.0-29b-a4b, Ornith: ornith-1.0-9b
Artificial Analysis Intelligence Index v4.3.2 ~23.6 63% Meta: muse-glimmer-30b, Qwen: Qwen3.5 9B, Xiaomi: MiMo-V2.5
GPQA ~75.3 63% DeepSeek: DeepSeek V4 Flash, DeepSeek: DeepSeek V4 Pro, NVIDIA Corporation: nvidia-nemotron-3-nano-30b-a3b
SuperGPQA ~59.46 63% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Qwen: qwen3.6-27b
Artificial Analysis Agentic Index (Maximum Reasoning) ~50.22 63% DeepSeek: DeepSeek V4 Pro, MiniMax: minimax-m2.7, Meta: Muse Spark
AIME 2025 ~91.64 62% Z.ai: GLM 5, LG AI Research: k-exaone-236b-a23b, webai-official: twil-8b
Humanity's Last Exam (with tools) ~33.77 62% NVIDIA Corporation: nvidia-nemotron-3-nano-30b-a3b, Meta: Muse Spark, Ornith: ornith-1.5-9b
MMLU-Redux ~90.34 62% Qwen: Qwen3.5 27B, Qwen: Qwen3.5 9B, Prism ML: ternary-bonsai-2-27b
Artificial Analysis Intelligence Index v4.1.1 (Maximum Reasoning) ~34.12 62% Meta: muse-glimmer-30b, Qwen: qwen3.6-27b, Google: Gemma 4 31B Instruct
CowQA ~91.77 61% Qwen: Qwen3.5 9B, DeepSeek: DeepSeek V4 Flash, Xiaomi: MiMo-V2.5
GSM8K ~99.6 60% DeepSeek: DeepSeek V4 Pro, Prism ML: ternary-bonsai-2-27b, Prism ML: ternary-bonsai-27b
LLMArena (llmarena.ru) (Snapshot from 19 june 2026) ~1078.91 60% OpenAI: gpt-oss-120b, DeepSeek: deepseek-v3.2, Meta Llama: Llama 3.3 70B Instruct
AA-Omniscience Accuracy ~21.54 59% NVIDIA Corporation: nvidia-nemotron-3.5-bolt-74b-a9b-24gb, ifm: k2-horizon-mova-36b-a4b, Xiaomi: mimo-v2-pro
HumanEval ~94.65 58% webai-official: twil-8b, Qwen: Qwen3 4B, Tencent: youtu-llm-2b
MBPP ~93.7 58% webai-official: twil-8b, Qwen: Qwen3 4B, Tencent: youtu-llm-2b