batch-api

💰 Optimize your Claude API usage to save 50-90% on costs with batching techniques and efficient request management.

EN ↔ 中文 i18n diff + Claude translate + HITL apply. Structured outputs + Batch API (50% off). pip · MCP server · MIT.

Local mock server for testing LLM retries, fallbacks, and provider failure handling without spending tokens.

WordPress Core Pre-Auth RCE via REST Batch Route Confusion + SQLi (CVE-2026-63030 + CVE-2026-60137)

loom-batch: run datasets of prompts (CSV, JSON, Parquet) through OpenAI, Anthropic, Gemini, OpenRouter and Qwen. Batch APIs at half price, or live. CLI + Python library.

Model router and eval lab: send each request to the cheapest Claude model and effort level that can handle it, and prove it with evals.

LangGraph pipeline that finds YouTube creators, scores them with Claude, and sends personalized outreach emails. Fully configurable via YAML.

Sanitized case study: LLM enrichment of 500K+ legal records with plain-English summaries, intent tags, and search metadata — dry-run gated, schema-validated, and routed through the provider Batch API for ~50% cost reduction.

Reconcile asynchronous LLM batch JSONL artifacts by custom_id