past.dev launches long-term memory API for AI agents, tops BEAM public benchmark
thisiskp_ · x · 2026-10-16
Developer thisiskp, who runs 10+ bots, recommends past.dev's long-term memory API for fixing agents that confidently cite outdated facts: querying a budget returns the current figure, a note that it changed in July, and the email that changed it.
Key details:
- Long-term, multiplayer memory layer for agents; #1 on BEAM, the largest public memory benchmark (92.08% at 100K tokens, 90.65% at 1M, 85.03% at 10M)
- Two core endpoints: /ingest stores text with original timestamps (backfilled data keeps its date), /recall returns ranked evidence with dated sources per identity
- Permission-aware so each person only sees what they're allowed to; SOC 2 Type II, ISO 27001/27701; managed or self-hosted
- Installs into an existing agent stack with a single prompt
More from coding & agent
- Running a multi-model UI design arena: Paseo orchestrates four design drafts per model — vista8 · 2026-10-16
- Cortex A/B-tests coding-agent skills and rules by replaying developer corrections before keeping them — Samuele Giampieri · 2026-10-16
- Tool calls are the hidden bottleneck killing AI coding speed — Crabbox and 5 fixes — steipete · 2026-10-16
- Agentic sandbox WorldBox doubles GPT-6 Astra's MazeBench score, 23 to 61 gems — ZhitingHu · 2026-10-16
- Two prompts: Gemini 4 Argon turns a single-player game into full multiplayer — chongdashu · 2026-10-16
- Anthropic and Gray Swan launch $300,000 indirect prompt injection red-teaming contest — bcherny · 2026-10-16