fix(zh): slice entity text from original input to avoid BERT tokenizer spaces

feat: smarter fallback — heuristic sufficiency + result merging (v3.1)

feat: dual-model routing + fallback strategy (v3.0)

feat: optional labels, bilingual auto-expansion, language hint (v2.0)

refactor: remove httpx proxy, call GLiNER model directly

fix: add /api/v1 prefix and restore port 4000

feat: GLiNER NER HTTP API