ENC-Bench - Coordinate Localization (Geographic): leaderboard

Metric: Acc@200px accuracy (%): share of ENC-Bench spatial reasoning answers with a marked point located by its latitude and longitude, converted to pixels, within 200 pixels of the target, numeric answers on ENC-Bench's 840 NOAA S-57 electronic navigational charts rendered with OpenCPN in day, dusk and night palettes at six scales, zero-shot; higher is better. Source: arxiv.org. Saturation forecast: Around 2028. 10 models tracked.

Top models

#ModelScoreOverall rank
1Gemini 2.5 Pro17.36#145
2Gemini 2.5 Flash16.77#237
3InternVL3-38B13.19#395
4GPT-4o12.2#333
5Qwen 3 VL 235B A22B Instruct11.31#264
6Qwen 3 VL 235B A22B (Thinking)11.01#228 (Qwen 3 VL 235B A22B)
7GLM-4.5V9.52#339
8Llama 4 Maverick Instruct9.13#439
9Qwen 3 VL 32B (Thinking)6.94#287 (Qwen 3 VL 32B)
10Qwen 3 VL 32B Instruct5.26#276

No result here: #3 Claude Opus 5.5, #5 GPT-6 Astra, #8 Claude Fable 5.1.

Interactive version: theaggregate.ai/benchmark?slug=enc-bench-coordinate-localization-geographic · How It Works · Data refreshed daily, snapshot 2026-10-11.