Copy-paste recipes for shipping production LLM apps. Every example runs against the live NexToken API with nex-pro as the default — Singapore-hosted, $0.10/1M input, 32K context.
Switch your existing OpenAI SDK code to NexToken by changing only api_key and base_url. Save ~95%.
End-to-end retrieval-augmented Q&A in Chinese using nex-embed-zh + nex-pro. Pgvector + chunking.
EN ↔ ZH ↔ JA ↔ KO translation with quality scoring. Streaming + tool calling for term-glossary injection.
12 MIN · LIVEClassify 100K rows of support tickets at 30% off retail via the batch endpoint. Includes resume + retry.
GitHub Actions workflow that runs nex-pro on every PR diff, comments inline. ~$0.02 per medium PR.
DRAFTChunk + summarise + risk-flag long contracts using nex-reasoning. Map-reduce pattern with citations.
Build a terminal chat UX that streams tokens. Hook stdin pipes to nex chat --stream.
Set spend caps per app, alert via Slack on 80%/100%, query MTD by provider. Production cost-control patterns.