SKT A.X K2 Card - BrowseComp (10 Searches Max): leaderboard
BrowseComp web-browsing questions with each model limited to at most 10 searches, run by SK Telecom. SKT A.X K2 model card benchmark table (July 2026), all models in thinking mode; accuracy in percent, higher is better. A.X K1 has no score.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Around May 2028. 7 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GLM-5.1 (Thinking) | 29.1 | #109 (GLM-5.1) |
| 2 | Qwen 3.5 397B A17B (Reasoning) | 26.9 | #154 (Qwen 3.5 397B A17B) |
| 3 | Kimi K2.6 (Thinking) | 21.5 | #113 (Kimi K2.6) |
| 4 | DeepSeek V4 Flash (Reasoning) | 16.8 | #155 (DeepSeek V4 Flash) |
| 5 | MiniMax-M2.7 | 14.2 | #270 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=skt-a-x-k2-card-browsecomp-10-searches-max · How It Works · Data refreshed daily, snapshot 2026-11-05.