Show HN: Halv cut AI agent cost by 57.1% using Jev
Halv/Blog/Inside Halv’s 57.1% cost reduction: the SWE-rebench dataTechnical report · Astra × JEV × SWE-rebenchInside Halv’s 57.1% cost reduction: the SWE-rebench dataAcross 14 tasks and three repetitions each, Halv and vanilla Codex both passed 25 of 42 runs. Halv recorded $144.67 in model cost versus $337.50.September 30, 2026· Pedro VillacaQuick answerHalv recorded 57.1% lower model cost across 42 paired repetitions, with 25 verifier passes per arm. The strongest task result was ArcadeDB-4455: 81.7% lower cost and 3/3 passes on both sides.
- ▪Halv/Blog/Inside Halv’s 57.1% cost reduction: the SWE-rebench dataTechnical report · Astra × JEV × SWE-rebenchInside Halv’s 57.1% cost reduction: the SWE-rebench dataAcross 14 tasks and three repetitions each, Halv and vanilla Codex both pa
- ▪Halv recorded $144.67 in model cost versus $337.50.September 30, 2026· Pedro VillacaQuick answerHalv recorded 57.1% lower model cost across 42 paired repetitions, with 25 verifier passes per arm.
- ▪The strongest task result was ArcadeDB-4455: 81.7% lower cost and 3/3 passes on both sides.
Hacker News (AI / LLM) files mainly under ai. We currently carry 7,079 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 | Halv |
| Canonical URL | https://halv.ai/blog/halv-swe-rebench-astra-42-pairs/ |
| Publication time | Wed, 30 Sep 2026 19:16:09 +0000 |
| Retrieval time | 2026-09-30T19:27:01.682Z |
| Last seen | 2026-09-30T19:27:01.682Z |
| 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 | yqx56IVHEkSA · 1 stories |
| 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
Halv/Blog/Inside Halv’s 57.1% cost reduction: the SWE-rebench dataTechnical report · Astra × JEV × SWE-rebenchInside Halv’s 57.1% cost reduction: the SWE-rebench dataAcross 14 tasks and three repetitions each, Halv and vanilla Codex both passed 25 of 42 runs. Halv recorded $144.67 in model cost versus $337.50.September 30, 2026· Pedro VillacaQuick answerHalv recorded 57.1% lower model cost across 42 paired repetitions, with 25 verifier passes per arm. The strongest task result was ArcadeDB-4455: 81.7% lower cost and 3/3 passes on both sides. This interim workflow comparison used more tokens with Halv.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Halv.
Discussion
0 commentsMore from Halv
