PCB-Bench - QA Routing SBERT: leaderboard
Task-2 open-ended routing question answering scored by SBERT sentence-embedding similarity.
Metric: SBERT similarity (%). Source: digailab.github.io. Saturation forecast: Not forecast. 9 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 Nano | 63.47 | #425 |
| 2 | Gemini 2.5 Flash Lite | 63.12 | #422 |
| 3 | Gemini 2.5 Pro | 63.11 | #152 |
| 4 | InternVL3-78B | 62.42 | #352 |
| 5 | Llama 4 Maverick | 61.08 | #464 |
| 6 | GPT-5 | 60.55 | #98 |
| 7 | Qwen 3 VL 235B A22B (Thinking) | 60.37 | #238 (Qwen 3 VL 235B A22B) |
| 8 | Qwen 3 VL 8B Instruct | 59.97 | #420 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=pcb-bench-qa-routing-sbert · How It Works · Data refreshed daily, snapshot 2026-10-20.