6G-Bench - AI-Native Networking and Agentic Control: leaderboard
Metric: Accuracy (%) on the AI-native networking and agentic control tasks (T11, T17-T20, T27) of 6G-Bench's 3,722 expert-validated four-option multiple-choice questions on network-level semantic reasoning for AI-native 6G networks, deterministic single-shot answers (temperature 0, one letter in a JSON object), group score is the unweighted mean of its task accuracies; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Llama 4 Maverick | 85.5 | #451 |
| 2 | GPT-5.2 Codex | 84 | #89 |
| 3 | Qwen3 Coder Next | 83.8 | #321 |
| 4 | GPT-5.2 Instant | 83.8 | #205 |
| 5 | Olmo 3.1 32B Instruct | 81.1 | #754 |
| 6 | DeepSeek V3.2 Exp | 80.7 | #227 |
| 7 | Qwen 3 235B A22B 2507 Instruct | 79.9 | #291 |
| 8 | DeepSeek V3.2 | 79.8 | #198 |
| 9 | Phi-4 | 79.5 | #701 |
| 10 | GPT-4o Mini | 79.3 | #588 |
| 11 | Hermes 4 70B | 79.3 | #489 |
| 12 | Claude Haiku 4.5 | 79 | #271 |
| 13 | Ministral 3 14B | 78.7 | #636 |
| 14 | Qwen 3 VL 32B Instruct | 78.1 | #276 |
| 15 | GPT-5 Mini | 77.5 | #176 |
No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.
Interactive version: theaggregate.ai/benchmark?slug=6g-bench-ai-native-networking-and-agentic-control · How It Works · Data refreshed daily, snapshot 2026-10-11.