Guides
One guide per kind of product, and one per agent framework. Each shows what memory changes for your users, how to map your product onto workspaces, peers and sessions, working code (TypeScript and Python, or the framework's own language, all of it run against a real Recall server on every change), the settings that matter, what it costs and how fast it is, and how to forget someone.
New to Recall? Start with the quickstart and Designing workspaces, peers and sessions. Building on an agent framework? The framework guides add Recall to the Vercel AI SDK, LangChain, LangChain.js, Mastra, the OpenAI Agents SDK or CrewAI in a few lines.
Use cases
Personal assistants and companionsAn assistant that knows the user's life, plans and preferences from the first message.
Customer supportAgents and bots that know each customer's history across tickets and channels.
Coding agentsClaude Code, Codex and your own agents that remember conventions and decisions.
Game NPCsCharacters that remember the player, keep their own point of view and gossip.
Sales and CRM copilotsA copilot that remembers every prospect, call and objection across the deal.
Multi-agent systemsAgents that share what they learn, each with its own view and its own key.
Tutoring and educationA tutor that tracks what each learner knows, mixes up and is working towards.
Discord and Slack botsA bot in a busy channel that keeps every member apart and remembers each one.
Health, fitness and coachingA coach that remembers goals, setbacks and what worked, handled with care.
Voice agentsPhone and voice assistants that know the caller, within a real-time budget.
Shopping assistantsRecommendations that remember sizes, tastes, budgets and what was returned.
Workplace assistantsAn assistant for a team that remembers projects, decisions and who owns what.
Switch to Recall
Build well
Designing workspaces, peers and sessionsHow to map your product onto Recall so memory stays separate where it should.
Reasoning levels and modelsWhen to read context, when to ask, and which reasoning level to pick.
Privacy, consent and forgettingWhat to store, how to tell users, and how to forget a person completely.
Cost and speedWhat a user costs per month and how fast each call is, with the arithmetic.
Frameworks
Vercel AI SDKMemory for generateText, streamText and agents: one middleware and three tools.
LangChain.jsA chat history in Recall sessions, a retriever over what Recall knows, and tools.
MastraOne processor for memory in every run, plus tools the agent can call.
LangChain and LangGraphMiddleware, tools and a retriever that give a LangChain agent memory of each user.
OpenAI Agents SDKA session kept in Recall, memory tools and instructions that know the user.
CrewAITools and task context so a crew remembers the people it works for.