TelcoAgent-Bench - Blueprint Reliability (English): leaderboard
TelcoAgent-Bench (2026) tests small language models as telecom troubleshooting agents: from the first message of an engineer the agent recognizes the intent, calls network tools along a reference path among distractor tools and writes a resolution summary. Blueprint reliability from 0 to 100, combining exact and near gold-path rates with stability across dialogues of one blueprint, in English; higher is better.
Metric: Blueprint reliability score x100 (0-100) on TelcoAgent-Bench (15 telecom troubleshooting intents, 49 blueprints with 30 sampled dialogues each; the agent starts from the engineer's first message and calls core and distractor network tools), English dialogues: a weighted sum of the exact gold-path rate, the gold-path rate within one edit, and one minus the pairwise edit-distance dispersion across a blueprint's samples (weights not printed; every cell fits 0.5, 0.3 and 0.2); higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 8 models tracked. Updated 7 Oct 2026.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Granite 3.3 8B Instruct | 46.8 | #3866 |
| 2 | Qwen 3 8B | 31.4 | #1873 |
| 3 | Mistral 7B Instruct | 26.3 | #7568 |
| 4 | Gemma 3 4B (IT) | 21.4 | #4139 |
| 5 | Llama 3 8B Instruct | 21.1 | #4852 |
| 6 | Qwen 2.5 7B Instruct | 18.9 | #2723 |
| 7 | Qwen-7B Chat | 14.7 | #7138 |
| 8 | granite-3.1-3B-a800m-instruct | 8.4 | #8600 |
No result here: #1 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=telcoagent-bench-blueprint-reliability-english · How It Works · Data refreshed daily, snapshot 2026-11-21.