{"slug": "an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all", "title": "An OpenAI flagship lost 38% of its daily usage in three days — then set three straight all-time highs", "summary": "Whatstrending.ai, a service that tracks daily request counts for AI models via OpenRouter's usage feed, reports that OpenAI's flagship model gpt-5.6-luna experienced a 38.3% drop in daily usage over three days, followed by three consecutive all-time highs within the same week. The data, stored append-only in Cloudflare D1, reveals that monthly aggregates mask significant day-level volatility in model usage.", "body_md": "Everyone publishes AI model leaderboards. Almost nobody publishes the **day-level curve** — and the\n\ncurve is where the actual story is.\n\nAt [whatstrending.ai](https://whatstrending.ai/models) we snapshot every model's daily request count\n\nfrom OpenRouter's usage feed into a database, once a day, `INSERT OR IGNORE`\n\n— a row can never be\n\nrevised after the fact. We're 78 days in. Here is what `openai/gpt-5.6-luna`\n\ndid over the last two\n\nweeks:\n\n| Date | Requests/day |\n|---|---|\n| Aug 04 | 162.9M |\n| Aug 05 | 180.8M |\n| Aug 06 | 193.8M |\n| Aug 07 | 210.8M |\n| Aug 08 |\n222.5M ← peak |\n| Aug 09 | 209.6M |\n| Aug 10 | 171.6M |\n| Aug 11 |\n137.3M ← floor: −38.3% off peak in 3 days |\n| Aug 12 | 143.2M |\n| Aug 13 | 157.3M |\n| Aug 14 |\n231.8M ← new all-time high, +47.4% in one day |\n| Aug 15 | 237.3M |\n| Aug 16 |\n243.5M ← third consecutive all-time high |\n\nA flagship model shed **more than a third of its daily traffic in three days**, then won all of it\n\nback — and more — inside the same week. Three consecutive days above the old peak as of Aug 16.\n\nThree checks, run before writing this:\n\n`deepseek/deepseek-v4-flash-0731`\n\nand `tencent/hy3`\n\nwere setting their own highs on the same days — so this is not a provider-wide\ntrailing-edge reporting droop.Monthly aggregates — which is what every public ranking shows — would have flattened this entire\n\nevent into \"usage grew\". The day-level series shows real flagship usage is far more volatile than\n\nanything the leaderboards suggest: a −38% drawdown and a full recovery to new highs fit inside eight\n\ndays. If you're making routing, capacity, or \"which model is winning\" decisions off monthly numbers,\n\nyou're averaging away the actual signal.\n\nThe live, daily-updated version of this curve is here:\n\n[https://whatstrending.ai/models](https://whatstrending.ai/models)\n\n*Method: daily snapshot of OpenRouter per-day request counts, 100 models/day, 78 days and counting,\nstored append-only ( INSERT OR IGNORE) in Cloudflare D1.*", "url": "https://wpnews.pro/news/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all", "canonical_source": "https://dev.to/codelong888/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-straight-all-time-4ebb", "published_at": "2026-08-16 01:41:04+00:00", "updated_at": "2026-08-16 02:11:08.708703+00:00", "lang": "en", "topics": ["ai-products", "ai-infrastructure", "developer-tools"], "entities": ["whatstrending.ai", "OpenRouter", "OpenAI", "gpt-5.6-luna", "Cloudflare D1", "deepseek/deepseek-v4-flash-0731", "tencent/hy3"], "alternates": {"html": "https://wpnews.pro/news/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all", "markdown": "https://wpnews.pro/news/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all.md", "text": "https://wpnews.pro/news/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all.txt", "jsonld": "https://wpnews.pro/news/an-openai-flagship-lost-38-of-its-daily-usage-in-three-days-then-set-three-all.jsonld"}}