--- title: Scaling Test-Time Compute in Search Mode | daily.dev description: Weaviate has released an `effort` parameter for the Query Agent's Search Mode in `weaviate-agents 1.8.0` and `agents-typescript-client 1.7.0`. Three tiers —... canonical: https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Scaling Test-Time Compute in Search Mode | daily.dev og:description: Weaviate has released an `effort` parameter for the Query Agent's Search Mode in `weaviate-agents 1.8.0` and `agents-typescript-client 1.7.0`. Three tiers —... og:url: https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti og:image: https://api.daily.dev/og/posts/cTVTTD5TI.png og:image:alt: Scaling Test-Time Compute in Search Mode og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Scaling Test-Time Compute in Search Mode **[Weaviate](https://daily.dev/sources/weaviate)** · 8 min read · 2 upvotes · 2 comments ## Summary Weaviate has released an `effort` parameter for the Query Agent's Search Mode in `weaviate-agents 1.8.0` and `agents-typescript-client 1.7.0`. Three tiers — medium, high, and ultrahigh — control how much compute is spent on query decomposition and reranking at inference time. Benchmarks across 8 datasets (BRIGHT, IRPAPERS, WixQA, OBLIQ-Bench) show consistent accuracy gains over Hybrid Search at every tier, with the largest improvements on reasoning-intensive retrieval tasks. On BRIGHT Biology, ultrahigh effort lifts nDCG@10 from 13.0 to 57.5 — roughly a 4x improvement. On simpler domain-specific benchmarks like WixQA, medium effort already captures most of the gain. The tradeoff is latency and cost, making tier selection a per-request decision based on application needs. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What does the new effort parameter in Weaviate's Query Agent Search Mode do? It lets developers choose between medium, high, and ultrahigh compute tiers per search request, trading latency and cost for retrieval accuracy. Higher effort scales computation at query decomposition and reranking. It is available in weaviate-agents 1.8.0 (Python) and agents-typescript-client 1.7.0 (JS/TS), and benchmarks show ultrahigh effort lifting nDCG@10 on BRIGHT Biology from 13.0 to 57.5 versus Hybrid Search. _daily.dev surfaces retrieval engineering updates like this for teams tuning search latency versus accuracy tradeoffs._ ### How much better is Weaviate Search Mode than Hybrid Search on hard reasoning retrieval benchmarks? On BRIGHT Biology, ultrahigh effort Search Mode reaches an nDCG@10 of 57.5 compared to 13.0 for Hybrid Search alone, roughly a fourfold improvement. On OBLIQ-Bench, ultrahigh effort achieves 24.4 Success@1 versus Hybrid Search hitting the floor, surfacing gold passages in fewer than 1 in 12 queries. Gains are largest on reasoning-intensive, long queries. _Developers comparing retrieval strategies track benchmark results like these on daily.dev before picking a search architecture._ ### When should I use medium effort instead of high or ultrahigh effort in Weaviate Search Mode? Medium effort is recommended for latency-sensitive or domain-specific but less reasoning-intensive tasks, since on benchmarks like IRPAPERS and WixQA it already captures most of the accuracy gain over Hybrid Search, with higher tiers adding only small increments. On WixQA, medium effort returns the correct top-ranked document for about two out of three queries. Reserve higher tiers for harder, reasoning-intensive workloads. _Teams weighing latency against accuracy in search pipelines follow guidance like this on daily.dev._ ## Community discussion Top comments from developers on daily.dev. **@agustinbarrientos** · 0 upvotes > Ultrahigh belongs in the retry path, not the default. **@kartiknvj** · 0 upvotes > That nDCG@10 jump from 13 to 57.5 on BRIGHT Biology is a huge spread, though it makes me want the latency and cost numbers right next to it, since ultrahigh effort is presumably not free. An exposed effort dial is a nice pattern because it lets you spend compute only on the hard queries instead of flat-rating everything. Did the gains hold on easier datasets, or was BRIGHT the one where decomposition paid off most? ## Similar posts on daily.dev - [Weaviate 1.37 Release](https://daily.dev/posts/weaviate-1-37-release-nnexx0ui6) · Weaviate · 7 upvotes · 0 comments - [Your agent wants to search like a 2010 quant](https://daily.dev/posts/your-agent-wants-to-search-like-a-2010-quant-lx3kju83b) · Vespa Blog · 1 upvotes · 0 comments - [3x Faster Search: Parallel Test-Time Scaling with Instructed-Retriever-1](https://daily.dev/posts/3x-faster-search-parallel-test-time-scaling-with-instructed-retriever-1-mzfvguqym) · databricks · 0 upvotes · 0 comments - [Measuring What Matters with Jules](https://daily.dev/posts/measuring-what-matters-with-jules-bwyrmpwn7) · Google Developers · 1 upvotes · 0 comments - [Evaluate agent performance](https://daily.dev/posts/evaluate-agent-performance-znjiurbqt) · Google Cloud · 1 upvotes · 0 comments --- Tags: [#rag](https://daily.dev/tags/rag), [#vector-search](https://daily.dev/tags/vector-search), [#weaviate](https://daily.dev/tags/weaviate) [View this post on daily.dev](https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Scaling Test-Time Compute in Search Mode","url":"https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti"},"datePublished":"2026-08-11T15:14:43.053Z","dateModified":"2026-09-13T21:22:18.055Z","description":"Weaviate has released an `effort` parameter for the Query Agent's Search Mode in `weaviate-agents 1.8.0` and `agents-typescript-client 1.7.0`. Three tiers —...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/73d591ec8ea1c4abd76ef17e55b448e0?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/73d591ec8ea1c4abd76ef17e55b448e0?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Weaviate","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Weaviate","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/07d0692410934a578c4d59e02f7ea685","url":"https://daily.dev/sources/weaviate"},"commentCount":2,"discussionUrl":"https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":2},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":2}],"keywords":"rag,vector-search,weaviate","timeRequired":"PT8M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Weaviate","item":"https://daily.dev/sources/weaviate"},{"@type":"ListItem","position":3,"name":"Scaling Test-Time Compute in Search Mode"}]} {"@context":"https://schema.org","@type":"WebPage","@id":"https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti","comment":[{"@type":"Comment","text":"Ultrahigh belongs in the retry path, not the default.","datePublished":"2026-08-12T05:02:38.281Z","url":"https://daily.dev/posts/cTVTTD5TI#c-RiFCv1Uaf","author":{"@type":"Person","name":"Agustin Barrientos","url":"https://daily.dev/agustinbarrientos","image":"https://media.daily.dev/image/upload/s--5ayxQnqn--/f_auto/v1788281802/avatars/avatar_wQYYVe5Tbj0NJ7C7qPoa8?_a=BAMAMicg0"}},{"@type":"Comment","text":"That nDCG@10 jump from 13 to 57.5 on BRIGHT Biology is a huge spread, though it makes me want the latency and cost numbers right next to it, since ultrahigh effort is presumably not free. An exposed effort dial is a nice pattern because it lets you spend compute only on the hard queries instead of flat-rating everything. Did the gains hold on easier datasets, or was BRIGHT the one where decomposition paid off most?","datePublished":"2026-08-11T16:57:35.946Z","url":"https://daily.dev/posts/cTVTTD5TI#c-ywfCD9rip","author":{"@type":"Person","name":"kartik-nvjk","url":"https://daily.dev/kartiknvj","image":"https://media.daily.dev/image/upload/s--3gGgsVCw--/f_auto/v1781456774/avatars/avatar_TvTVeiMdkRCqWUDullFmy?_a=BAMAMiWQ0"}}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/scaling-test-time-compute-in-search-mode-ctvttd5ti#faq","mainEntity":[{"@type":"Question","name":"What does the new effort parameter in Weaviate's Query Agent Search Mode do?","acceptedAnswer":{"@type":"Answer","text":"It lets developers choose between medium, high, and ultrahigh compute tiers per search request, trading latency and cost for retrieval accuracy. Higher effort scales computation at query decomposition and reranking. It is available in weaviate-agents 1.8.0 (Python) and agents-typescript-client 1.7.0 (JS/TS), and benchmarks show ultrahigh effort lifting nDCG@10 on BRIGHT Biology from 13.0 to 57.5 versus Hybrid Search. daily.dev surfaces retrieval engineering updates like this for teams tuning search latency versus accuracy tradeoffs."}},{"@type":"Question","name":"How much better is Weaviate Search Mode than Hybrid Search on hard reasoning retrieval benchmarks?","acceptedAnswer":{"@type":"Answer","text":"On BRIGHT Biology, ultrahigh effort Search Mode reaches an nDCG@10 of 57.5 compared to 13.0 for Hybrid Search alone, roughly a fourfold improvement. On OBLIQ-Bench, ultrahigh effort achieves 24.4 Success@1 versus Hybrid Search hitting the floor, surfacing gold passages in fewer than 1 in 12 queries. Gains are largest on reasoning-intensive, long queries. Developers comparing retrieval strategies track benchmark results like these on daily.dev before picking a search architecture."}},{"@type":"Question","name":"When should I use medium effort instead of high or ultrahigh effort in Weaviate Search Mode?","acceptedAnswer":{"@type":"Answer","text":"Medium effort is recommended for latency-sensitive or domain-specific but less reasoning-intensive tasks, since on benchmarks like IRPAPERS and WixQA it already captures most of the accuracy gain over Hybrid Search, with higher tiers adding only small increments. On WixQA, medium effort returns the correct top-ranked document for about two out of three queries. Reserve higher tiers for harder, reasoning-intensive workloads. Teams weighing latency against accuracy in search pipelines follow guidance like this on daily.dev."}}]} ```