AINewsnow

A half-million-token prompt found both needles on my 3090

This story is from 2026-10-19. It is preserved in the archive; the latest stories are on the live feed.

Every stack I've seen sells the context window as a model spec, and every fix for oversize is RAG, and what I measured on my own stack this month says the ceiling you actually feel lives in a config line. Mine was 131072. The model held retrieval exact at 509,084 tokens in one request. The lie I fo…

Read the full story at DEV Community — AI ↗

Timeline · 1 report

  1. 2026-10-19 22:16 · DEV Community — AI
    A half-million-token prompt found both needles on my 3090

More stories

  1. MiniMax launches M3.1 Flash, confirming Space Bunny was an early build — Vercel Blog
  2. Introducing Clairic, a new System One class of extraction models — Cloudflare Blog — AI
  3. OpenAI Staffer Charged With Bugging Office Stays Jailed for Now — Bloomberg AI
  4. How HP Is Building an AI-Ready CDP with Databricks CustomerLake — Databricks Blog
  5. Structured outputs for Anthropic Claude models on Amazon Bedrock — AWS Machine Learning Blog
  6. Musk pledges 1 SpaceX share for every child in America — Axios AI+
  7. Developers can now access SageMaker Studio through the new AWS experience — AWS Machine Learning Blog
  8. Nscale, Anthropic in Talks for $45 Billion Cloud Computing Deal — Bloomberg AI

Get the daily brief of stories like this at 6:30 every morning →