Tutorials
Price Comparison Agent for Google, Amazon and TikTok Shop
Build a price comparison agent API workflow: Google product cards, Amazon and TikTok Shop listings, model-number matching in Python, GPT-6 Luna for edge cases.
ReadSandBase Blog — Page 7
Insights on AI agents, model routing, and building production-ready AI systems.
Page 7
Tutorials
Product Review Analysis Agent for TikTok Shop and Xiaohongshu
Build a product review analysis agent for TikTok Shop and Xiaohongshu, then check its labels against 40 hand-labeled reviews: GPT-6 Luna vs GPT-6.1 Sol.
Read
Tutorials
Public Opinion Agent: Weibo, Zhihu, Bilibili on GPT-6 Luna
Build a public opinion monitoring agent for Weibo, Zhihu and Bilibili, then test GPT-6 Luna vs GPT-6.1 Sol on 50 hand-labeled posts for accuracy and cost.
Read
Model Comparison
Best Search API for Research Agents: 6 Tools Tested (2026)
Best search API for research agents, measured: Exa, Tavily, Firecrawl, Cloudsway, Scholar and Exa Answer with GPT-6.1 Sol on 20 fresh questions, 3 runs each.
Read
Tutorials
SandBase MCP in Claude Code: Setup and 12 Real Runs
Set up SandBase MCP in Claude Code with the official CLI, then see 12 headless runs on Xiaohongshu, Douyin and image tasks: tool calls, pass rate, cost.
Read
Developer Tools
Social Data API vs Web Scraping for AI Agents (2026)
Structured social data APIs vs web scraping for AI agents in 2026: reliability, legal exposure, cost, latency and coverage, plus a tested Douyin API example.
Read
Tutorials
Build a Social Listening Agent: Weibo + Douyin
Tutorial: an agent that checks Weibo hot search and the Douyin hot account billboard every 15 minutes, flags brand mentions and posts Slack alerts.
Read
Best Of
Top 5 Social Media Data APIs for AI Agents (2026)
A practical guide to social media data APIs that AI agents can use in 2026, covering Chinese and Western platforms, pricing, and agent-readiness.
Read
Model Comparison
LLM Agent Tool Calling Benchmark 2026: 12 Models Tested
LLM agent tool calling benchmark 2026: 12 models, 17 scored tasks, 3 runs each. Accuracy, cost per task, latency and why most misses were hand arithmetic.
Read