{"schema_version":"onlylabs.public_signal.v1","title":"Cloudflare (Workers AI) Writing: Cutting our inference bill by 68% by being obsessive about optimization","description":"Cloudflare (Workers AI) writing signal with public source context, captured evidence pages, related signals, and category-scoped analysis context.","url":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6","json_url":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6/signal.json","generated_at":"2026-10-25T18:28:36.128Z","evidence_latest_fetched_at":"2026-10-19T20:02:57.080313+00:00","signal_first_seen_at":"2026-10-19T20:00:52.918964+00:00","org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud","category_label":"Neocloud","dossier_url":"https://onlylabs.fyi/labs/cloudflare","dossier_json_url":"https://onlylabs.fyi/labs/cloudflare/dossier.json"},"related_urls":{"signal":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6","signal_json":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6/signal.json","source":"https://blog.cloudflare.com/obsessive-inference-optimization/","lab_dossier":"https://onlylabs.fyi/labs/cloudflare","lab_dossier_json":"https://onlylabs.fyi/labs/cloudflare/dossier.json","analysis":"https://onlylabs.fyi/analysis/cloudflare","analysis_json":"https://onlylabs.fyi/analysis/cloudflare/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/cloudflare/evidence.json","category":"https://onlylabs.fyi/neoclouds","category_json":"https://onlylabs.fyi/neoclouds.json","category_feed":"https://onlylabs.fyi/neoclouds/feed.xml","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","topic":"https://onlylabs.fyi/topics/talking","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neocloud","data_business":null},"answer_pack":{"answer":"Cloudflare (Workers AI) published Cutting our inference bill by 68% by being obsessive about optimization. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Substantive engineering post on inference cost optimization. · Cutting our inference bill by 68% by being obsessive about optimization | Cloudflare Blog Skip to content When we previously blogged about our AI code reviewer in April.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","signal_desk":"talking","source_context":{"source_url":"https://blog.cloudflare.com/obsessive-inference-optimization/","source_host":"blog.cloudflare.com","occurred_at":"2026-10-19T19:12:53+00:00","first_seen_at":"2026-10-19T20:00:52.918964+00:00","date_source":"rss.item_date","context":null},"context_markers":[{"label":"Lab","value":"Cloudflare (Workers AI)","source":"signal"},{"label":"Signal desk","value":"talking","source":"signal"},{"label":"Source host","value":"blog.cloudflare.com","source":"source"},{"label":"Author","value":"Ryan Skidmore","source":"source"},{"label":"Notability","value":"Substantive engineering post on inference cost optimization.","source":"signal"},{"label":"Watch term","value":"Eval methodology","source":"evidence"},{"label":"Watch term","value":"Infrastructure","source":"evidence"},{"label":"Watch term","value":"Safety and alignment","source":"evidence"},{"label":"Watch term","value":"Agents and tool use","source":"evidence"}],"evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://blog.cloudflare.com/obsessive-inference-optimization/"],"related_signals":6,"has_source_url":true,"latest_page_fetched_at":"2026-10-19T20:02:57.080313+00:00"},"data_business":{"matches":false,"lanes":[],"matched_terms":[],"score":null,"reason":null},"agent_handoff":{"signal_json":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6/signal.json","dossier_json":"https://onlylabs.fyi/labs/cloudflare/dossier.json","analysis_json":"https://onlylabs.fyi/analysis/cloudflare/analysis.json","analysis_evidence_json":"https://onlylabs.fyi/analysis/cloudflare/evidence.json","topic_signals_json":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","topic_feed":"https://onlylabs.fyi/topics/talking/feed.xml?category=neocloud","category_signals_json":"https://onlylabs.fyi/signals.json?category=neocloud","data_radar_json":null,"opportunities_json":null},"analysis_playbook":{"objective":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","evidence_focus":["post title","source URL","captured page text","HN traction","linked model or paper references","publication date"],"extraction_questions":["Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Which writing reframes a recent release, model, hiring wave, or policy stance?","Which posts mention data, evals, infrastructure, safety, or deployment workflows?"],"signal_questions":["What public theme, launch framing, or research direction does this writing signal expose?","Which themes are labs choosing to explain publicly?","Which posts are attracting outside discussion?","Do the 6 related writing signals show a repeated pattern?"],"output_fields":["org","theme","public_framing","traction","evidence_url"],"data_business_relevance":"Data-business lane extraction is scoped to frontier labs; for this category, keep conclusions tied to category-specific strategy, source evidence, and follow-up questions.","required_sources":[{"label":"signal_json","url":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6/signal.json","required":true},{"label":"source","url":"https://blog.cloudflare.com/obsessive-inference-optimization/","required":true},{"label":"dossier_json","url":"https://onlylabs.fyi/labs/cloudflare/dossier.json","required":true},{"label":"analysis_evidence_json","url":"https://onlylabs.fyi/analysis/cloudflare/evidence.json","required":true},{"label":"topic_signals_json","url":"https://onlylabs.fyi/topics/talking/signals.json?category=neocloud","required":false},{"label":"data_radar_json","url":null,"required":false}],"expected_output":["one-paragraph source-grounded interpretation","category-specific implication","confidence and missing evidence","recommended next source to inspect"],"prompt_seed":"Using only the linked onlylabs JSON, captured source context, and cited evidence, analyze Cloudflare (Workers AI)'s writing signal \"Cutting our inference bill by 68% by being obsessive about optimization\" for neocloud strategy."},"semantic_triples":[{"subject":"Cloudflare (Workers AI)","predicate":"published","object":"Cutting our inference bill by 68% by being obsessive about optimization","text":"Cloudflare (Workers AI) published Cutting our inference bill by 68% by being obsessive about optimization."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"is classified as","object":"writing signal","text":"Cutting our inference bill by 68% by being obsessive about optimization is classified as writing signal."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"belongs to","object":"talking desk","text":"Cutting our inference bill by 68% by being obsessive about optimization belongs to talking desk."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Cutting our inference bill by 68% by being obsessive about optimization has evidence coverage 1 captured evidence page."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has captured page count","object":"1","text":"Cutting our inference bill by 68% by being obsessive about optimization has captured page count 1."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has readable page count","object":"1","text":"Cutting our inference bill by 68% by being obsessive about optimization has readable page count 1."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has related signal count","object":"6","text":"Cutting our inference bill by 68% by being obsessive about optimization has related signal count 6."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has analysis playbook objective","object":"Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.","text":"Cutting our inference bill by 68% by being obsessive about optimization has analysis playbook objective Turn public writing and discussion into a readable map of research themes, product framing, policy posture, launch narratives, and market attention.."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has source host","object":"blog.cloudflare.com","text":"Cutting our inference bill by 68% by being obsessive about optimization has source host blog.cloudflare.com."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has lab","object":"Cloudflare (Workers AI)","text":"Cutting our inference bill by 68% by being obsessive about optimization has lab Cloudflare (Workers AI)."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has signal desk","object":"talking","text":"Cutting our inference bill by 68% by being obsessive about optimization has signal desk talking."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has source host","object":"blog.cloudflare.com","text":"Cutting our inference bill by 68% by being obsessive about optimization has source host blog.cloudflare.com."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has author","object":"Ryan Skidmore","text":"Cutting our inference bill by 68% by being obsessive about optimization has author Ryan Skidmore."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has notability","object":"Substantive engineering post on inference cost optimization.","text":"Cutting our inference bill by 68% by being obsessive about optimization has notability Substantive engineering post on inference cost optimization.."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has watch term","object":"Eval methodology","text":"Cutting our inference bill by 68% by being obsessive about optimization has watch term Eval methodology."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has watch term","object":"Infrastructure","text":"Cutting our inference bill by 68% by being obsessive about optimization has watch term Infrastructure."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has watch term","object":"Safety and alignment","text":"Cutting our inference bill by 68% by being obsessive about optimization has watch term Safety and alignment."}]},"intelligence":{"signal_desk":"talking","answer":"Cloudflare (Workers AI) published Cutting our inference bill by 68% by being obsessive about optimization. This talking signal gives public context for research themes, product direction, policy, or launch framing. High-signal details: Substantive engineering post on inference cost optimization. · Cutting our inference bill by 68% by being obsessive about optimization | Cloudflare Blog Skip to content When we previously blogged about our AI code reviewer in April.... onlylabs links this event to 1 captured evidence page and 6 related writing signals.","semantic_triples":[{"subject":"Cloudflare (Workers AI)","predicate":"published","object":"Cutting our inference bill by 68% by being obsessive about optimization","text":"Cloudflare (Workers AI) published Cutting our inference bill by 68% by being obsessive about optimization."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"is classified as","object":"writing signal","text":"Cutting our inference bill by 68% by being obsessive about optimization is classified as writing signal."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"belongs to","object":"talking desk","text":"Cutting our inference bill by 68% by being obsessive about optimization belongs to talking desk."},{"subject":"Cutting our inference bill by 68% by being obsessive about optimization","predicate":"has evidence coverage","object":"1 captured evidence page","text":"Cutting our inference bill by 68% by being obsessive about optimization has evidence coverage 1 captured evidence page."}]},"signal":{"id":"7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6","url":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6","json_url":"https://onlylabs.fyi/signals/7eb513b4-24b8-4b24-bf6e-4af46eaf5eb6/signal.json","source_url":"https://blog.cloudflare.com/obsessive-inference-optimization/","title":"Cutting our inference bill by 68% by being obsessive about optimization","summary":"Cloudflare (Workers AI) published a writing signal. onlylabs watches public writing for research themes, product direction, and model-launch context.","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-19T19:12:53+00:00","first_seen_at":"2026-10-19T20:00:52.918964+00:00","date_source":"rss.item_date","evidence_coverage":{"target_pages":1,"captured_pages":1,"readable_pages":1,"capture_methods":["plain"],"missing_page_urls":[],"failed_page_urls":[],"blocked_page_urls":[],"page_urls":["https://blog.cloudflare.com/obsessive-inference-optimization/"]},"facets":{},"traction":{"github_stars":null,"hn_points":null,"hn_comments":null,"hn_story_id":null,"hf_downloads":null,"hf_likes":null},"data_radar":null},"primary_evidence_page":{"is_primary":true,"source_match":true,"url":"https://blog.cloudflare.com/obsessive-inference-optimization/","final_url":"https://blog.cloudflare.com/obsessive-inference-optimization/","title":"Cutting our inference bill by 68% by being obsessive about optimization","http_status":200,"content_type":"text/html","capture_method":"plain","fetched_at":"2026-10-19T20:02:57.080313+00:00","bytes":317699,"raw_path":"68e2ee5d21f7dc1bf4b7c7b6d9eebb47bc8199714c890d52200121be88c48fd3.html","content_hash":"f61421cdc481d11c477ba72e8f202042ba31cd6154434cd1a31f470adab1a192","excerpt_chars":1200,"truncated":true,"excerpt":"Cutting our inference bill by 68% by being obsessive about optimization | Cloudflare Blog Skip to content When we previously blogged about our AI code reviewer in April , an average review cost a palatable $1.19. By the end of August, that figure had shot up to about $2.46. We hadn't made any significant architectural changes: the coordinator was still handing work to the same set of specialist reviewers, and the same risk tiers decided how intensive of a review we gave. Around mid-September we tasked our agents with working out where those tokens were going and cutting them — aggressively. With code quality at high effort, the average review now costs just $0.77, down 68% from the $2.43 it averaged over our two most expensive weeks before we started. Our main goal was to cut costs, but as part of this work we also managed to optimize the reviewer, so that the median review now takes just 1m 32s instead of the 4m 09s it took before. About 40% of the savings came from changes and optimizations to the reviewer's code, and the rest from experimenting with which models we use and how much effort they put in. How a review doubled in cost The chart below shows the cost of an average..."},"evidence_pages":[],"related_signals":[{"id":"9c9d357b-b379-4de0-ab2c-30764ba08632","url":"https://onlylabs.fyi/signals/9c9d357b-b379-4de0-ab2c-30764ba08632","source_url":"https://blog.cloudflare.com/origin-range-requests/","title":"Origin Range Requests: Serving large files without range-request tuning","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-23T18:22:59+00:00","first_seen_at":"2026-10-23T20:00:52.925659+00:00","date_source":"rss.item_date"},{"id":"36b8c4da-4f96-4abc-8a89-7ff3e47942cd","url":"https://onlylabs.fyi/signals/36b8c4da-4f96-4abc-8a89-7ff3e47942cd","source_url":"https://blog.cloudflare.com/post-quantum-geo-key-manager/","title":"Bringing post-quantum cryptography to geographically fenced key management","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-23T18:06:56+00:00","first_seen_at":"2026-10-23T20:00:52.925659+00:00","date_source":"rss.item_date"},{"id":"d9e0cfd4-b38f-4c15-bd9e-81cfe0ac8dfb","url":"https://onlylabs.fyi/signals/d9e0cfd4-b38f-4c15-bd9e-81cfe0ac8dfb","source_url":"https://blog.cloudflare.com/k-isms/","title":"Localizing compliance at scale: achieving South Korea’s ISMS certification","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-22T13:00:00+00:00","first_seen_at":"2026-10-22T16:00:54.670606+00:00","date_source":"rss.item_date"},{"id":"c52bb1e0-df36-4ab7-b8cd-a48a6f37dfef","url":"https://onlylabs.fyi/signals/c52bb1e0-df36-4ab7-b8cd-a48a6f37dfef","source_url":"https://blog.cloudflare.com/full-transaction-waf/","title":"Introducing full transaction detection for Cloudflare WAF","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-21T14:10:42+00:00","first_seen_at":"2026-10-21T16:00:52.275864+00:00","date_source":"rss.item_date"},{"id":"6c59f359-43de-4c9d-be79-876752c91605","url":"https://onlylabs.fyi/signals/6c59f359-43de-4c9d-be79-876752c91605","source_url":"https://blog.cloudflare.com/php-vulnerabilities-waf-protection/","title":"Cloudflare deploys protections against php2shell vulnerability","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-21T13:56:45+00:00","first_seen_at":"2026-10-21T16:00:52.275864+00:00","date_source":"rss.item_date"},{"id":"7cf512fd-f264-4ff0-ba38-acd212a3e47e","url":"https://onlylabs.fyi/signals/7cf512fd-f264-4ff0-ba38-acd212a3e47e","source_url":"https://blog.cloudflare.com/webmcp-commerce/","title":"How Commerce.com and Cloudflare are getting storefronts ready for AI agents","context":null,"kind":{"key":"post_published","label":"Writing"},"org":{"slug":"cloudflare","name":"Cloudflare (Workers AI)","category":"neocloud"},"occurred_at":"2026-10-21T13:00:00+00:00","first_seen_at":"2026-10-21T16:00:52.275864+00:00","date_source":"rss.item_date"}]}