Skip to content
Models and systems measured
182
Scores recorded
185
Sources
4
IQ weight
1.41

Score distribution

38.2%
Mean
35.8%
Median
23.1%
Std dev
7.9%
10th percentile
78.2%
90th percentile
2.9%
Min
2.9%91%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.41 for chat models, measured over 176 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#Model or systemScore
101Granite 3.0 8B Instruct· Hugging Face Open LLM Leaderboard data31.6%
102NousResearch: Hermes 2 Pro - Llama-3 8B· Hugging Face Open LLM Leaderboard data30.7%
103Nous-Hermes-2-Mixtral-8x7B-SFT· Hugging Face Open LLM Leaderboard data30.6%
104ERNIE 4.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim30.4%
105Neural Daredevil 8B abliterated· Hugging Face Open LLM Leaderboard data30.3%
106Mixtral 8X7b V0.1· Hugging Face Open LLM Leaderboard data30.3%
107Qwen1.5-14B· Hugging Face Open LLM Leaderboard data30.1%
108Mixtral 8x7B Instruct v0.1· Hugging Face Open LLM Leaderboard data29.7%
109Mistral Nemo· Hugging Face Open LLM Leaderboard data29.7%
110Llama 3.1 8B Instruct· Hugging Face Open LLM Leaderboard data29.4%
111mistral-nemo-base-2407· Hugging Face Open LLM Leaderboard data29.4%
112OpenHermes-2-Mistral-7B· Hugging Face Open LLM Leaderboard data29.3%
113Qwen2.5 Coder 7B Instruct· Hugging Face Open LLM Leaderboard data28.9%
114Aya Expanse 8B· Hugging Face Open LLM Leaderboard data28.5%
115Qwen2.5-Coder-7B· Hugging Face Open LLM Leaderboard data28.4%
116Llama 3 8B Instruct· Hugging Face Open LLM Leaderboard data28.2%
117OpenHermes 2.5 Mistral 7B· Hugging Face Open LLM Leaderboard data27.8%
118granite-3-1-8b-base· Hugging Face Open LLM Leaderboard data26%
119Mistral-7B-OpenOrca· Hugging Face Open LLM Leaderboard data25.8%
120Qwen2.5 3B Instruct· Hugging Face Open LLM Leaderboard data25.8%
121Mistral (7B) Instruct v0.3· Hugging Face Open LLM Leaderboard data25.6%
122Llama 3.1 8B· Hugging Face Open LLM Leaderboard data25.3%
123GLM-4-9b-chat· Hugging Face Open LLM Leaderboard data25.2%
124Meta Llama 3 8b· Hugging Face Open LLM Leaderboard data24.5%
125Llama 3.2 3B· Hugging Face Open LLM Leaderboard data24.1%
126OpenChat 3.5· Hugging Face Open LLM Leaderboard data24%
127mistral-7b-v0.3· Hugging Face Open LLM Leaderboard data24%
128Qwen1.5-7B· Hugging Face Open LLM Leaderboard data23.1%
129Mistral 7B Instruct v0.2· Hugging Face Open LLM Leaderboard data22.9%
130fimbulvetr-11b-v2· Hugging Face Open LLM Leaderboard data22.7%
131Qwen1.5-7B-Chat· Hugging Face Open LLM Leaderboard data22.4%
132Granite 3.1 2B Instruct· Hugging Face Open LLM Leaderboard data21.8%
133Granite 3.0 2B Instruct· Hugging Face Open LLM Leaderboard data21.7%
134Qwen2.5 Math 7B Instruct· Hugging Face Open LLM Leaderboard data21.5%
135Zephyr 7B Beta· Hugging Face Open LLM Leaderboard data21.5%
136Starcoder2 15b· Hugging Face Open LLM Leaderboard data20.4%
137Yi-6B200k· Hugging Face Open LLM Leaderboard data20.1%
138Qwen2.5 1.5B Instruct· Hugging Face Open LLM Leaderboard data19.8%
139Gemma 2 2B IT· Hugging Face Open LLM Leaderboard data18%
140DeepSeek R1 Distill Qwen 32B· Hugging Face Open LLM Leaderboard data17.1%
141Qwen2.5 1.5B· Hugging Face Open LLM Leaderboard data16.7%
142Qwen1.5-4B-Chat· Hugging Face Open LLM Leaderboard data16.3%
143Qwen1.5-4B· Hugging Face Open LLM Leaderboard data16.2%
144Gemma 1.1 7B IT· Hugging Face Open LLM Leaderboard data15.9%
145NexusRaven-V2-13B· Hugging Face Open LLM Leaderboard data15.3%
146vicuna-7b-v1.5· Hugging Face Open LLM Leaderboard data15.2%
147Nous-Hermes-llama-2-7b· Hugging Face Open LLM Leaderboard data13.8%
148Qwen2 1.5b Instruct· Hugging Face Open LLM Leaderboard data13.7%
149granite-3.0-3b-a800m-instruct· Hugging Face Open LLM Leaderboard data13.2%
150Qwen2.5 Math 1.5b Instruct· Hugging Face Open LLM Leaderboard data12.9%

Showing 101–150 of 179 results.

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
DeepSeek V4.1 Flash BaseEM · 3-shot· Reported by DeepSeek· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.86.1%
MiMo V2 Flash Baseofficial Base table · 3-shot· Reported by Xiaomi MiMo· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.88.5%
Qwen3.5-397B-A17B-BaseQwen3.5 release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.91%
Qwen3.7-Plus-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.89.4%
Qwen3.8-27B-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.89.6%
Qwen3.8-Flash-Next-BaseQwen3.8-Flash-Next release/model card · base-model table· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.90.9%

Data credits: Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · Data by LLM Stats · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.