--- title: "Search · style controlled: model rankings and source evidence · Undominated.ai" canonical: https://undominated.ai/rankings/arena-search-style-control/ description: "Search · style controlled ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence." --- # Search · style controlled: model rankings and source evidence · Undominated.ai > Search · style controlled ranked by Arena score. Inspect exact model variants, source dates, measurement limits and downloadable evidence. [All rankings](/rankings/) / Search and documents # Search · style controlled Human preference ratings for the overall category of this Arena task. Scores from different boards are not interchangeable. **42 entries** · Arena score · Higher is better Source publication: 2026-10-13. Retrieved 2026-11-05. Source: [Arena (lmarena-ai)](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset) · [CC-BY-4.0](https://creativecommons.org/licenses/by/4.0/). Ranks use confidence bounds. Several models can share a rank; the rows within a shared rank are displayed by score. Search · style controlled · Arena score | Significance rank | Model / evaluated variant | Arena score | 95% confidence interval | Votes | | --- | --- | --- | --- | --- | | 1 | gpt-5.6-sol-xhigh openai | 1,257.93 | 1,251.46 – 1,264.4 | 60,437 | | 1 | gpt-6.1-sol-max openai | 1,249.92 | 1,240.9 – 1,258.94 | 5,555 | | 1 | gpt-6-astra-max openai | 1,246.24 | 1,238.2 – 1,254.28 | 18,483 | | 3 | claude-fable-5.1-high anthropic | 1,233.37 | 1,226.2 – 1,240.54 | 12,782 | | 2 | qwen3.8-max alibaba | 1,229.5 | 1,216.2 – 1,242.79 | 1,870 | | 4 | claude-fable-5-high anthropic | 1,229.02 | 1,222.44 – 1,235.6 | 73,596 | | 4 | gpt-6-sol-max openai | 1,228.21 | 1,220.77 – 1,235.65 | 11,313 | | 4 | gpt-5.5-search openai | 1,226.45 | 1,221.46 – 1,231.45 | 125,957 | | 4 | claude-opus-4-6-search anthropic | 1,222.05 | 1,217.36 – 1,226.73 | 173,191 | | 7 | claude-opus-5-high anthropic | 1,214.35 | 1,207.46 – 1,221.24 | 13,285 | | 8 | claude-opus-5.5-medium anthropic | 1,211.23 | 1,204.13 – 1,218.32 | 10,842 | | 10 | claude-opus-4-7 anthropic | 1,209.71 | 1,204.77 – 1,214.64 | 127,506 | | 9 | gpt-6-luna-max openai | 1,209.25 | 1,201.86 – 1,216.65 | 10,793 | | 10 | gemini-3.1-pro-grounding google | 1,206.62 | 1,201.83 – 1,211.4 | 151,436 | | 10 | grok-4.20-multi-agent-beta-0309 xai | 1,206.38 | 1,200.83 – 1,211.94 | 126,238 | | 10 | grok-4.5 xai | 1,206.19 | 1,200.29 – 1,212.09 | 76,458 | | 11 | claude-sonnet-4-6-search anthropic | 1,201.93 | 1,197.26 – 1,206.6 | 173,986 | | 11 | gemini-3-pro-grounding google | 1,200.75 | 1,195.26 – 1,206.24 | 38,857 | | 11 | grok-4.20-beta1 xai | 1,199.31 | 1,193.25 – 1,205.38 | 57,049 | | 13 | gpt-5.4-search openai | 1,196.26 | 1,190.59 – 1,201.94 | 126,730 | | 13 | o3-search openai | 1,196.16 | 1,189.43 – 1,202.9 | 20,770 | | 15 | claude-opus-4-8 anthropic | 1,195.25 | 1,189.61 – 1,200.89 | 106,304 | | 17 | grok-4-1-fast-search xai | 1,195 | 1,189.92 – 1,200.08 | 85,919 | | 17 | claude-sonnet-5-search anthropic | 1,193.16 | 1,186.87 – 1,199.45 | 72,978 | | 17 | grok-4.3 xai | 1,192.79 | 1,187.93 – 1,197.65 | 128,393 | | 15 | ernie-5.1 baidu | 1,192.14 | 1,182.66 – 1,201.63 | 4,024 | | 20 | gemini-3-flash-grounding google | 1,188.68 | 1,184.28 – 1,193.07 | 196,399 | | 19 | gpt-5-search openai | 1,187.73 | 1,180.35 – 1,195.11 | 20,865 | | 20 | claude-opus-4-5-search anthropic | 1,186.69 | 1,180.96 – 1,192.42 | 65,047 | | 24 | gpt-5.1-search openai | 1,183.64 | 1,178.27 – 1,189.02 | 63,042 | | 28 | claude-sonnet-4-5-search anthropic | 1,175.98 | 1,170.74 – 1,181.21 | 144,153 | | 31 | claude-opus-4-1-search anthropic | 1,167.09 | 1,162.11 – 1,172.07 | 79,806 | | 32 | gpt-5.2-search openai | 1,161.4 | 1,155.6 – 1,167.2 | 55,634 | | 32 | grok-4-fast-search xai | 1,161.08 | 1,156.26 – 1,165.89 | 43,083 | | 33 | claude-opus-4-search anthropic | 1,151.84 | 1,146.29 – 1,157.39 | 31,506 | | 35 | gemini-2.5-pro-grounding google | 1,144.23 | 1,139.43 – 1,149.03 | 86,689 | | 35 | ppl-sonar-pro-high perplexity | 1,143.22 | 1,137.32 – 1,149.11 | 28,899 | | 36 | ppl-sonar-reasoning-pro-high perplexity | 1,138.5 | 1,132.57 – 1,144.43 | 29,439 | | 36 | gpt-5.2-search-non-reasoning openai | 1,135.76 | 1,129.95 – 1,141.56 | 79,796 | | 40 | grok-4-search xai | 1,122.74 | 1,116.55 – 1,128.92 | 19,195 | | 41 | api-gpt-4o-search openai | 1,083.31 | 1,071.88 – 1,094.73 | 3,399 | | 42 | diffbot-small-xl diffbot | 1,062.55 | 1,053.84 – 1,071.27 | 6,368 | ## What this ranking can tell you - The population is the models in this source board, including variants and models without an available API price. - Ranks use the published confidence bounds: one plus the number of models whose lower bound exceeds this model’s upper bound. Shared ranks are not proof of equal performance. ## Check the source The original evidence is retained byte for byte. The downloadable board includes the source URLs, revision and SHA-256 hashes. [Download this board and provenance →](/data/rankings/arena-search-style-control.json) Source files and licences - [metadata · a7a159f557a1](/data/rankings/evidence/a7a159f557a1e594a5d41218fe8d53c63c63daa39b655e51d6fc5b482a17d059.json) · [upstream source ↗](https://huggingface.co/api/datasets/lmarena-ai/leaderboard-dataset) - [licence · 88a46d472f35](/data/rankings/evidence/88a46d472f35274fb6b7a852c3d16c8ee8007d51b37f290161537bec26895ee4.txt) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/raw/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/README.md) - [rows · 2ac3d9a63ca9](/data/rankings/evidence/2ac3d9a63ca97bc5d8feecdd100b5de42e16a1d5d219d82e96c7d637eb69556f.parquet) · [upstream source ↗](https://huggingface.co/datasets/lmarena-ai/leaderboard-dataset/resolve/8b6deb69ace01ee8dcda9e46eba7a5e8a808ec53/search_style_control/latest-00000-of-00001.parquet) [Explore other model types →](/rankings/)[Read the benchmark guide →](/benchmarks/) ## Continue your investigation - [Understand the measurements](/benchmarks/) - [Compare task performance](/best/) - [Compare embedding models](/embeddings/) - [Inspect ranking rules](/methodology/)