Skip to content

Benchmark

Arena ELO - Instruction Following

Instruction following benchmarks measure how closely models follow instructions. Higher scores are better.

Models measured
366
Scores recorded
423
Sources
1
IQ weight
1.18

Score distribution

1,294.5
Mean
1,332.7
Median
155.5
Std dev
1,056.7
10th percentile
1,465
90th percentile
780.7
Min
780.71,539.1
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.18 for chat models, measured over 366 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
151Qwen3.5 27B· LMArena data1,392.6
152Qwen3 Next 80B A3B Instruct· LMArena data1,388.7
153DeepSeek V3.1· LMArena data1,388.4
154GPT 5High· LMArena data1,388.2
155Grok 4 FastNon-Reasoning· LMArena data1,386.8
156Grok 40709· LMArena data1,386.6
157nova-experimental-chat12-10· LMArena data1,386.2
158nova-experimental-chat11-10· LMArena data1,386.1
159Qwen 3 235B A22B2507, thinking· LMArena data1,386.1
160Step 3.5 Flash· LMArena data1,385
161Grok 4 FastReasoning· LMArena data1,382.2
162DeepSeek R10528· LMArena data1,381.3
163GPT-5 Chat· LMArena data1,380.9
164DeepSeek V3.1 Terminus· LMArena data1,380.8
165Qwen3.5 35B A3B· LMArena data1,379.9
166nova-experimental-chat26-01-10· LMArena data1,379.9
167Gemini 3.1 Flash Lite· LMArena data1,378.1
168GPT-5.3 Chat· LMArena data1,377.7
169Grok 4.1 FastReasoning· LMArena data1,376.4
170Claude Sonnet 4Thinking 32K, 20250514· LMArena data1,375.9
171Hunyuan T120250711· LMArena data1,375.3
172Solar Pro 4· LMArena data1,375.1
173Qwen3.5 Flash· LMArena data1,375
174Qwen3 VL 235Bthinking· LMArena data1,373.8
175Claude Opus 420250514· LMArena data1,373.4
176MiMo V2 FlashThinking· LMArena data1,373.3
177Muse Glimmer· LMArena data1,372.9
178OpenAI o32025-04-16· LMArena data1,367.7
179o12024-12-17· LMArena data1,367.2
180GPT 4.12025-04-14· LMArena data1,366.9
181Gemini 2.5 Flash-Lite PreviewThinking, 06-17· LMArena data1,366.4
182Grok 4.3· LMArena data1,365.6
183nova-experimental-chat10-20· LMArena data1,365.5
184Qwen3 30B A3B Instruct2507· LMArena data1,362.8
185Qwen 3 235B A22BNon-Thinking· LMArena data1,361.9
186GPT 5.4 NanoHigh· LMArena data1,361.3
187DeepSeek R1· LMArena data1,357.3
188GPT-5 miniHigh· LMArena data1,356.8
189Grok 3 MiniHigh· LMArena data1,356.5
190Qwen3 Coder· LMArena data1,355.1
191Gemini 2.5 Flash-Lite PreviewNon-Thinking, 09-2025· LMArena data1,354.9
192GLM 4.5 Air· LMArena data1,353.4
193MiniMax M2.5· LMArena data1,352.9
194Claude Sonnet 420250514· LMArena data1,352
195Claude Sonnet 3.7Thinking 32K, 20250219· LMArena data1,351.7
196GLM 4.6v· LMArena data1,350.8
197Granite 4.2 30B· LMArena data1,348.4
198Nemotron 3 Super 120B A12B· LMArena data1,347.2
199Kimi K2 Preview0905· LMArena data1,347
200Grok 3 Minibeta· LMArena data1,346.3

Showing 151–200 of 423 results.

Data credits: LMArena, CC BY 4.0 (licence)