past.dev

past.dev is a memory layer for AI agents that maintains a record of current facts, including history and source citations for retrieved information.

past.dev acts as a memory layer for AI agents, replacing the need to resend entire conversation histories or rely on simple vector searches. It allows applications to ingest text, emails, call notes, and tickets, which the system then parses to identify facts, rules, and events stamped with specific times. When an agent requests context, it receives ranked facts along with the dated sources that verify them. The platform supports fine-grained access control, ensuring agents only retrieve information they are permitted to see. It is designed for developers building production-grade agents that require accurate, traceable, and time-aware recall.

past.dev at a glance

Best for: Software engineer, Researcher
Runs on: API
Pricing: Pay as you go (free credits to start, then metered); Flex from $99/month; Enterprise custom pricing.
Released: October 2026
Category: AI research tool
Website: past.dev

Key features

Ingestion of text, emails, transcripts, and tickets
Dated source citations for every retrieved fact
Automatic contradiction resolution and historical tracking
Access control based on audience and identity
Time-aware recall using custom query timestamps
API-based architecture for agent integration
SOC 2 Type II, ISO 27001, and ISO 27701 certification

Use it for

Updating customer account records based on recent communications
Retrieving the current status of project budgets or tickets
Resolving contradictory information across multiple chat sessions
Enabling agents to answer questions based on historical state

Alternatives to past.dev

Playground — Persists agent memory across sessions and searches by meaning.
Exa — Provides high-accuracy web search and deep research endpoints for AI agents.
Hindsight — It uses a four-way retrieval strategy and background consolidation to maintain accurate memory.
Memori — Reduces LLM token usage and spend by over 97% with targeted recall.
Supermemory — Integrates long-term memory into AI coding agents with a single prompt.
Tigerless Agent Memory — Maintains agent history in portable, greppable Markdown files without API keys.
Agentmemoryrepo — Enables agents to navigate memories like a wiki and share knowledge across sessions.
TencentDB Agent Memory — It unifies cross-agent memory assets, skill libraries, wikis, and code graphs.
AI Memory — Shared memory for coding agents that works across machines and team members.
ReMe — It enables local-first, file-native long-term memory sharing across multiple AI agents.
Maximem — It delivers fast retrieval and structured memory across frameworks.
Mem0 — It provides drop-in persistent memory for AI agents without changing pipelines.

All past.dev alternatives