TelcoAgent-Bench - Sequence Alignment (Arabic): 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. Sequence alignment from 0 to 100 between the agent tool calls and the gold tool path in Arabic dialogues; higher is better.
Metric: Sequence alignment 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), Arabic dialogues: per sample, the longest common subsequence of the agent's tool calls with the gold tool path over the gold length, times a penalty for calls beyond the gold length; 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 | 62.3 | #3756 |
| 2 | Qwen 3 8B | 60.3 | #1890 |
| 3 | Llama 3 8B Instruct | 47.2 | #4858 |
| 4 | Gemma 3 4B (IT) | 46.4 | #4227 |
| 5 | Mistral 7B Instruct | 43.6 | #7544 |
| 6 | Qwen 2.5 7B Instruct | 42.4 | #2732 |
| 7 | granite-3.1-3B-a800m-instruct | 33.4 | #8610 |
| 8 | Qwen-7B Chat | 30 | #7152 |
No result here: #2 Claude Fable 5.2, #3 Gemini 4 Neon, #4 GPT-6.2 Astra.
Interactive version: theaggregate.ai/benchmark?slug=telcoagent-bench-sequence-alignment-arabic · How It Works · Data refreshed daily, snapshot 2026-11-22.