Training SID-1 to beat GPT-5 at search with 1k+ QPS RL
SID-1 is a new search model designed to outperform existing models like GPT-5 in terms of recall and efficiency. It utilizes large-scale reinforcement learning to improve search latency and reduce costs significantly. The model's iterative approach allows it to adaptively gather context and provide better search results compared to traditional static retrieval methods.
- ▪SID-1 nearly doubles recall over classical retrieval pipelines.
- ▪It outperforms GPT-5 at significantly lower latency and cost.
- ▪The model was trained using large-scale reinforcement learning at over 1,000 searches per second.
Hacker News (Newest) files mainly under programming. We currently carry 5,306 of its stories.
Story provenance
Source · retrieval · rights · ranking — open for full record
inspect →
Story provenance
Attribution is not the same as permission. This drawer separates discovery metadata, excerpts, WeSearch-generated summaries, reuse status, and whether the publisher receives the visit. Nothing here claims a legal grant the publisher has not made.
Record
| Original publisher | turbopuffer |
| Canonical URL | https://turbopuffer.com/blog/reinforcement-learning-sid-ai |
| Publication time | Wed, 20 May 2026 19:31:15 +0000 |
| Retrieval time | 2026-05-20T19:35:02.977Z |
| Last seen | 2026-05-20T19:35:02.977Z |
| Headline source | Publisher (no WeSearch rewrite) |
| Excerpt source | publisher body |
| Excerpt method | First ~120 words (~800 chars) of extracted publisher body, fair-use limited. |
| Summary | WeSearch · cerebras-chat (WeSearch summarizer) |
| Summary source text | contentText |
| Citation coverage | Summary is a WeSearch-generated derivative; primary citation is the original publisher URL. |
| Cluster | svwMvjkAdDrv |
| Cluster logic | Grouped by semantic title/content similarity across sources within a rolling window. Same-publisher template collisions are excluded from coverage comparison. |
| Ranking reason | Story pages are not engagement-ranked. Hub feeds use recency, with optional source-diversified chronological ordering (cap consecutive stories per source). No personalized ranking. |
| Publisher visit | Yes — open original |
| Substitutes article? | No — link-out required for full text |
Rights status (four layers)
WeSearch handling by dimension
| Indexing | May the item be indexed (stored, ranked, made findable)? | Allowed |
| Snippet | May a short excerpt of the publisher's text be shown? | Allowed |
| AI summary | May WeSearch generate its own short summary of the article? | Limited |
| Retrieval / RAG | May the content be exposed for third-party retrieval-augmented generation? | Not asserted |
| Model training | May the content be used to train AI models? | Not asserted |
| Commercial reuse | May the content be reused commercially? | Not permitted |
Basis: Derived from the published RSS/Atom feed. Contact: [email protected]. Reviewed: 2026-07-24.
Opening excerpt (first ~120 words) tap to expand
Training SID-1 to beat GPT-5 at search with 1k+ QPS RLMay 20, 2026•Max Rumpf (Co-founder of SID), Sam Dauncey (Researcher at SID)guestGiven sufficient search tools and time, humans can find almost anything. We search, read results, adapt, and search again until we find the information we seek. We're Max and Sam, co-creators of SID-1, an agentic search model that builds upon this idea. As a result of its training, SID-1 nearly doubles recall over classical retrieval pipelines and outperforms frontier LLMs at orders of magnitude lower latency and cost.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at turbopuffer.
Discussion
0 commentsMore from turbopuffer
