Research — Maximem field notes
Original research, benchmarks, and architecture deep-dives on AI memory and agent context management.
Maximem Synap Updates: Higher Scores, 17 Integrations, and a Live Playground
Synap updates: 92% LongMemEval (up from 90.2%), 93.2% LOCOMO, 17 framework integrations, a browser playground, public pricing, and a free accuracy eval on your own agent.
May 27, 2026
Read More
How Synap Works Under the Hood
We launched Maximem Synap today. Here's a peek into how it is built.
April 11, 2026
Read More
Synap Scores 92% on LongMemEval, 93.2% on LoCoMo: What the Numbers Mean
Synap outperforms existing memory systems by redesigning context management on leading benchmarks; delivering higher accuracy, lower latency, and stable performance at scale through structured, domain-specific architectures.
April 10, 2026
Read More
File Search vs Vector Search for RAG: 50,000 Documents, 5,000 Queries, Full Results
We ran exact-match keyword search against vector search across five datasets, 50,000 documents and 5,000 queries, scored by exact document ID with no LLM judge. Vector search wins on average, and the entire margin comes from one dataset.
2026-01-15
Read More