Each example is a full, working artifact — hero, code, architecture. Fork the repo, plug in your sk-aig-… key, and ship in under an hour.
Spin up a five-agent swarm that runs locally on Hermes, routes every call through AIgateway, and automates your inbox, calendar, research, coding, and reporting — using Kimi K2.6 on the free tier until Apr 30.
Ship one MCP server on the AIgateway MCP surface and every agent — Claude Code, Cursor, Codex, Cline, anything that speaks MCP — can call its tools. No infra to host, no auth to wire, no schema-sync headaches.
A planner fans out parallel web searches, an extractor pulls citations, a contradiction check catches hallucinations, and a reporter writes a 1,000-word memo with sources — end-to-end on Kimi K2.6's free tier, swap any step to Opus when the stakes are high.
A ReAct loop with four tools — read, plan, edit, test — that reads your repo, makes multi-file patches, runs the tests, and iterates on failures. One AIgateway key, any model slug, full control. No vendor lock-in.
Three modalities, one key: Deepgram streaming transcription feeds Kimi K2.6, which streams into ElevenLabs — voice-to-voice with p50 latency under 650ms. Runs on the free Kimi tier through Apr 30.
Drop-in extraction pipeline: upload a PDF/invoice/receipt, let Kimi K2.6 or Opus 4.7 read it whole, enforce your schema with JSON mode, and get machine-readable data in 30 lines. 96% field-level accuracy vs OCR on the benchmark set inside.
Send the same 50-row dataset to Opus 4.7, GPT-5.4, Kimi K2.6, Gemini 3.1, and Llama 4.1 in parallel through one AIgateway key, grade every response with an LLM judge, and publish a scorecard — 40 lines of Python, no eval framework required.
A single OpenAI-SDK call that tries Claude Opus first, falls back to GPT-5.4 on rate limit or timeout, and finally to Kimi K2.6 — with retry budgets, per-provider timeouts, and header-level observability. Zero extra lines of orchestration.
A real 30-day rebuild: semantic cache, complexity-based routing, and per-feature hard caps took one team's GPT-only monthly spend from $4,800 to $1,420 without touching product code. Every lever is a header, a config, or a one-line POST.
We publish a new example whenever a builder asks for one twice. Email support@aigateway.sh with a one-line description — most land within 48 hours.