Exa 3.0 ties Perplexity at #1 on Artificial Analysis Search Index with score of 80
yoimnotkesku · x · 2026-10-20
Exa released Exa 3.0 with a new Dynamic Highlights mode that allocates a shared text budget across pages for a query. On the Artificial Analysis Search Index, Exa Search (auto, Dynamic Highlights) scores 80, tying Perplexity Search (medium) for #1.
Artificial Analysis benchmarked both modes using a Stirrup agent harness with GPT-5.6 Luna (medium): auto balances quality and speed, while the new lite mode targets lower cost and latency — scoring 79 at roughly $0.045/task, placing it in the most attractive quadrant on the cost-quality chart.
More from coding & agent
- uv author explains: env markers as decision diagrams, minimal equivalent DNF is computationally hard — charliermarsh · 2026-10-20
- uv's scaling pain: complex resolutions explode the intermediate DNF, clause comparison gets expensive — charliermarsh · 2026-10-20
- Yeufonic: open-source frontend for YuE2/ComfyUI aimed at musicians, with local LoRA training — AcrobaticMaize2408 · 2026-10-20
- OpenCode ships DeepSeek V4.1 Flash 1017 update, zero changes needed for users — intellectronica · 2026-10-20
- "The LLM is the brain, the harness is the body": framing agent engineering — EricBuess · 2026-10-20
- Coinbase lets AI agents trade perps and index products via x402 protocol — MurrLincoln · 2026-10-20