Same model, different provider, different structured output
A technical note discusses issues with structured extraction in an LLM workflow. The same model was routed to different providers, leading to inconsistent extraction results. Key lessons include the importance of logging provider identity and considering provider choice in structured-output workloads.
- ▪The extraction layer failed to capture budget and date flexibility from user inputs.
- ▪Initial theories about prompt specificity and schema shape did not resolve the extraction issues.
- ▪Removing assistant history improved extraction success, indicating that the shape of the assistant message affected outcomes.
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 | Guilherme Costa |
| Canonical URL | https://guilhermesfc.com/provider-variance-structured-extraction.html |
| Publication time | Mon, 25 May 2026 18:58:16 +0000 |
| Retrieval time | 2026-05-25T19:07:40.295Z |
| Last seen | 2026-05-25T19:07:40.295Z |
| 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 | MZgfegdUTDHT |
| 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
Model provider variance in structured extraction May 2026 · Technical note TL;DR We spent several hours debugging what looked like a prompt-engineering issue in a structured extraction pipeline. This turned out not to be a prompt issue. The same OpenRouter model (google/gemini-3-flash-preview) was being routed to different upstream providers, and one provider consistently failed to extract certain fields when assistant history contained list-shaped content. Practical lesson: log the routed provider treat provider identity as part of the request fingerprint be careful with latency-based routing on structured-output workloads For structured-output workloads, provider choice can be part of correctness, not just latency or cost.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Guilherme Costa.
Discussion
0 commentsMore from Guilherme Costa
