Artificial Analysis Launches Search API Benchmark Covering 26 Products Across 14 Providers
ArtificialAnlys · x · 2026-11-21
Artificial Analysis released the Search API Benchmark & Leaderboard, evaluating 26 Search API products from 14 providers to help developers choose web-search tools for AI agents.
- Scoring: equal-weighted mean of DeepSearchQA (F1), BrowseComp (exact-answer accuracy), and AA-Omniscience (accuracy) on a 0-100 scale, plus per-task cost and latency.
- Token usage: Parallel Search (advanced) uses 82k tokens per task, second-lowest of 26; the lowest is OpenAI Web Search (medium) at 41k; Parallel Search (fast) at 116k is comparable to Keenable Search (pro) at 118k.
- Open source: the agent harness used for benchmark runs, Stirrup, is open-sourced; providers can apply to be included.
- Data as of Nov 19, 2026, distinguishing integrated first-party search tools from external search harnesses.
More from coding & agent
- LangChain ships four updates: LangSmith Agent Platform, trace data controls, Tool Gateway — LangChain · 2026-11-21
- Grok Bot Gets Its Own Email Addresses to Sign Up and Handle Promos for You — elonmusk · 2026-11-21
- CapCut × ChatGPT plugin turns chat into a video generator and editor — SimplyAnnisa · 2026-11-21
- Princeton & Meta's Strategist evolves inference strategies, beating baselines on 24/30 benchmarks at 39% lower cost — 141_1337 · 2026-11-21
- MendAgent: a multi-agent self-healing system that auto-repairs Dockerized microservices — ExpressionOne8668 · 2026-11-21
- Dev builds daily X-digest Grok bot that auto-blocked 19 reply bots, heading to marketplace — petergyang · 2026-11-21