PulseAugur
EN
LIVE 16:05:46

Character.ai boosts chatbot memory and models for deeper conversations

Character.ai has launched significant upgrades to its platform, including new AI models and enhanced memory features for its chatbots. The platform is rolling out "Story Memory" and "Facts" to help characters retain context and backstory over longer conversations. These updates are powered by new chat styles, PipSqueak 2 for all users and improved DeepSqueak for premium subscribers, aiming to improve in-character consistency and writing quality. AI

IMPACT Enhances user experience by allowing AI characters to maintain context and personality over extended interactions, potentially increasing engagement.

RANK_REASON Product updates including new models and core feature enhancements for a popular AI chatbot platform.

Read on arXiv cs.IR (Information Retrieval) →

AI-generated summary · Google Gemini · from 3 sources. How we write summaries →

Character.ai boosts chatbot memory and models for deeper conversations

COVERAGE [3]

  1. arXiv cs.IR (Information Retrieval) TIER_1 English(EN) · Sankar Arumugam ·

    CALMem : Application-Layer Dual Memory for Conversational AI

    Large language models (LLMs) operate within fixed context windows that fundamentally limit conversational continuity. When context fills, compaction discards history irreversibly; when sessions end, all memory resets to zero. Existing solutions-larger context windows, retrieval-a…

  2. Character.ai blog TIER_1 English(EN) ·

    Smarter Memory for Smarter Chats

    Long chats, big plots, and hours of backstory. Your Characters now keep up with all of it.  Story Memory, Facts, and Memory Usage are rolling out across Character.ai, giving you new ways to shape what your Character knows and lock in the moments that matter. You

  3. Character.ai blog TIER_1 English(EN) ·

    April Update: New Model, Memory, and Lorebook

    Today I want to share some big updates. Model improvements, Memory upgrades, and a feature our community has been asking for since long before I joined. All of this was shaped by community feedback and testing, and I