GSMA Open-Telco - TeleMath: leaderboard
The "TeleMath" subtask of GSMA Open-Telco. A task from the GSMA Open-Telco LLM Benchmark, evaluating models on telecommunications domain knowledge, standards and network engineering problems.
Metric: Score (%). Source: huggingface.co. Saturation forecast: Around January 2027. 90 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 3 Flash (Preview) | 80 | #82 |
| 2 | GPT-5 | 79 | #98 |
| 3 | Gemini 3 Pro (Preview) | 78 | #66 |
| 4 | Kimi K2.5 | 76 | #144 |
| 5 | Gemini 2.5 Pro | 75 | #152 |
| 6 | Claude Opus 4.6 | 75 | #63 |
| 7 | O3 | 74 | #124 |
| 8 | Gemini 3.1 Pro (Preview) | 73 | #58 |
| 9 | Claude Opus 4.1 | 73 | #140 |
| 10 | Claude Opus 4.5 | 72 | #83 |
| 11 | O1 | 72 | #183 |
| 12 | GPT-5.2 | 71 | #110 |
| 13 | GPT-5 Mini | 71 | #186 |
| 14 | Claude Opus 4 | 71 | #163 |
| 15 | Claude Sonnet 4.6 | 70 | #90 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=gsma-open-telco-telemath · How It Works · Data refreshed daily, snapshot 2026-10-20.