Plastic Labs
Plastic Labs develops Honcho, an AI-native memory and reasoning platform powered by proprietary Neuromancer models that give AI applications and agents persistent user identity and context. It serves developers building AI apps and multi-agent systems via API and open-source distribution.
- Company typePrivate
- Founded2023
- HeadquartersNew York, United States
- Headcount1–10
- GTM typeB2B and B2C
- OfferingSoftware
What Plastic Labs does
Plastic Labs is an engineering-driven AI lab building at the intersection of machine learning and cognitive science. Its core product, Honcho, is an AI-native memory and reasoning infrastructure platform that treats memory as a dynamic reasoning task rather than static storage, enabling AI applications and agents to maintain persistent personal identity and context across conversations. The platform is powered by the company's proprietary Neuromancer reasoning model series, with Neuromancer XR (a fine-tuned Qwen3-8B) deployed for explicit and deductive reasoning and Neuromancer MR in training for meta-reasoning through induction and abduction. Architectural innovations include the Memory as Reasoning framework spanning four logical certainty levels, a Peer Paradigm that replaces the traditional user-assistant model with peer-to-peer identity for multi-agent systems, and agentic components such as the Dialectic retrieval agent and Dreaming background processor.
Plastic Labs serves AI application developers building agent-based and language model applications that require stateful memory, personalization, and multi-agent coordination. The primary customer segment is horizontal, targeting developers building adaptive agent memory, context management, personalization, and multi-agent systems. Distribution is entirely product-led and self-serve across four channels: a managed SaaS platform at honcho.dev, open-source Honcho on GitHub for self-hosting, the consumer-facing Honcho Chat assistant, and the x402-powered marketplace for AI agent creations. The company is headquartered in New York with 1-10 employees and was founded in 2023.
The business model is usage-based on the managed platform, charging $2 per million tokens ingested and per-query fees starting at $0.001 depending on reasoning level, with the get_context() method free and unlimited. Open-source users self-host without paying Plastic Labs. Additional monetization vectors in development include paid dreaming tasks once configuration controls are exposed. Plastic Labs raised a $5.35M pre-seed round in April 2025 led by Variant, White Star Capital, and Betaworks, with Mozilla Ventures, Greycroft, Seed Club Ventures, Differential Ventures, and angels participating, bringing total disclosed funding to approximately $5.95M across two rounds.
Plastic Labs firmographics
Firmographics- Name
- Plastic Labs
- Legal name
- Plastic Labs
- Website
- https://plasticlabs.ai
- Company type
- Private
- Founded year
- 2023
- Operating status
- Operating
- Headcount range
- 1–10 employees
- Short description
- Plastic Labs develops Honcho, an AI-native memory and reasoning platform powered by proprietary Neuromancer models that give AI applications and agents persistent user identity and context. It serves developers building AI apps and multi-agent systems via API and open-source distribution.
- Ownership category
- akta.pro rank
Plastic Labs industry classification
Industry- Product category
- AI Memory Infrastructure
- NAICS
- Software Publishers (5132), Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (51821)
- SIC
- Services-Prepackaged Software (7372)
- akta.pro primary industry
- End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management) (HDAEANAA)
- akta.pro secondary industries
- Model Development & Training Platforms (AutoML, Notebooks, Feature Stores) (HDAEANAB), Conversational AI Platforms (chat/voice bots, orchestration) (HDAAAFAC)
Keywords
Where Plastic Labs is headquartered
LocationHeadquarters
- HQ city
- New York
- HQ country
- United States
- HQ region
- North America
Offices1 record
Markets served
Plastic Labs business model
Business model- GTM type
- B2B and B2C
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations
Revenue model
- Honcho Platform (Managed SaaS): Token-based pricing model for the hosted Honcho platform. Charges for tokens ingested at $2 per million tokens. Per-query pricing for chat endpoint starting at $0.001 per query depending on reasoning level. get_context() method is unlimited and free.
- Open Source Software: Honcho is available as open-source software for self-hosting, with pricing managed by users themselves.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Usage-based | Pay-as-you-go | Token-based ingestion pricing at $2 per million tokens |
| Usage-based | Pay-as-you-go | Per-query chat endpoint pricing starting at $0.001 |
| Subscription | Monthly | Unlimited get_context() method |
| Freemium | Pay-as-you-go | Dreaming tasks (free during beta) |
Go-to-market motion2 records
Distribution channels4 records
Marketing channels5 records
Plastic Labs product offering
Product offeringCore offering
Plastic Labs builds Honcho, an AI-native memory and reasoning platform that creates self-improving peer representations of users, agents, and other entities by applying formal logical reasoning (explicit, deductive, inductive, abductive) over conversational data. The platform is delivered as a cloud-hosted API with usage-based pricing, an open-source self-host option, and consumer-facing demo applications (Honcho Chat, Penny for Your Thoughts, YouSim) that showcase personalized AI experiences and an x402-powered marketplace for AI creations.
Product overview
Plastic Labs offers a unified product portfolio centered on AI-native memory and personal identity reasoning. The core platform is Honcho, an AI-native memory solution powered by the Neuromancer reasoning model series. Honcho Chat, Penny for Your Thoughts, and YouSim are demo applications showcasing different use cases of the Honcho platform—from general AI assistance to expertise marketplaces to identity simulation. The architecture treats memory as a reasoning task rather than static storage, with Neuromancer XR (deployed) and Neuromancer MR (in training) providing multi-level logical reasoning capabilities for modeling personal identity.
Differentiator
Problem solved
Functional benefit
Brands
- Honcho: AI-native memory and reasoning infrastructure for apps and agents. A continual learning system for modeling personal identity.
- Neuromancer
- Honcho Chat
- Penny for Your Thoughts
- YouSim
- Xeno Grant
Products and services
- Honcho AI-native memory and reasoning infrastructure for apps and agents. A continual learning system that models personal identity by reasoning over user/agent data to create self-improving peer representations. Sold as a cloud-hosted API and as open-source software.
- Neuromancer Series of custom reasoning models dedicated to AI-native memory and social cognition. Neuromancer XR performs explicit and deductive reasoning; Neuromancer MR (in training) handles meta-reasoning for inductive and abductive conclusions.
- Honcho Chat Personalized AI assistant platform built on Honcho's memory infrastructure. Features state-of-the-art memory, custom identities, shareable artifacts and themes, and an x402-powered marketplace for selling AI creations. Targeted at consumer end-users.
- Penny for Your Thoughts Honcho and x402-powered personal expertise market where users share expertise via AI interviews and sell access to their knowledge via crypto micropayments. Enables agents to pay humans for premium contextual information.
- YouSim
Quantifiable outcome
- 86.9% accuracy on LoCoMo memory benchmark vs 69.6% base Qwen3-8B and 80.0% Claude 4 Sonnet
- +3 more outcomes
Companies that use Plastic Labs
Customer profileSegments2 records
Ideal customer profiles3 records
Plastic Labs technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration2 records
AI capability5 records
Feature6 records
Plastic Labs partnerships and signals
Strategic signalPartnerships
Two partnerships are on record, tiered core.
- Solana FoundationcoreCo-founder of Xeno Grant program, a $15,000 grant initiative for autonomous AI agents. Collaboration aims to support independent AI research, decentralization, and innovation at the edge.
- TextArenacoreCollaboration to develop SPIRAL, a reinforcement learning framework for training LLMs through self-play on text-based games. TextArena provides Python library for text-only games as RL environments.
Scale indicators5 records
Recent moves6 records
Expansion highlights6 records
Plastic Labs competitors and assessment
Company assessmentDirect peers
- Zep (Zep AI): Zep provides long-term memory and context management APIs for AI agents, with GraphRAG-based recall and summarization. Directly competes with Honcho in the agent memory infrastructure layer.
- Mem0: Mem0 offers a memory layer for AI applications and agents, providing personalized memory storage and retrieval. Competes head-to-head with Plastic Labs' Honcho in the developer-facing memory-as-a-service category.
- Letta (formerly MemGPT): Letta builds stateful agents with hierarchical memory, providing APIs for long-context AI applications. Directly comparable to Honcho's memory-as-reasoning thesis for AI agent developers.
- Cognee: Cognee provides an open-source knowledge and memory graph layer for AI applications using ECL pipelines and ontologies. Overlaps with Honcho's peer representation and reasoning-based memory approach.
- Supermemory: Supermemory offers an infinite-context memory API for LLMs, targeting the same developer pain point (context rot, long-term recall) that Honcho addresses.
Broad incumbents
- LangChain: LangChain is a broad AI orchestration framework with memory components integrated into its agent and chain abstractions. Larger, well-funded incumbent that competes with Plastic Labs at the application layer where memory is one feature among many.
- LlamaIndex: LlamaIndex provides a data framework for LLM applications including memory and chat engine components. Established incumbent in RAG and agent memory space with broader scope than Plastic Labs.
- Pinecone: Pinecone is a vector database widely used as memory infrastructure for AI applications. While not a direct competitor to reasoning-based memory, it represents an adjacent approach to solving long-term recall that some developers may choose instead.
Emerging players
- Inworld AI: Inworld AI provides memory and character persistence for AI agents and NPCs, focused on gaming and interactive experiences. Overlaps with Honcho's multi-agent and peer-centric architecture for dynamic AI relationships.
- Letta Cloud / MemGPT open source: Open-source memory layer for stateful agents with similar developer-focused, API-first positioning as Honcho. Competes for the same developer mindshare in the agentic AI memory category.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat5 records
Key risks6 records
Key highlights7 records
Customer concentration
Plastic Labs social profiles
Digital presencePlastic Labs financial estimates
Financial estimateRevenue estimate
Valuation estimate
Plastic Labs leadership team
Management profileNumber of profiles
Profiles8 records
Plastic Labs funding detail
Funding detailFunding overview
Funding rounds4 records
Investors7 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Plastic Labs M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Plastic Labs
What does Plastic Labs do?
Plastic Labs builds Honcho, an AI-native memory and reasoning platform that creates self-improving peer representations of users, agents, and other entities by applying formal logical reasoning (explicit, deductive, inductive, abductive) over conversational data. The platform is delivered as a cloud-hosted API with usage-based pricing, an open-source self-host option, and consumer-facing demo applications (Honcho Chat, Penny for Your Thoughts, YouSim) that showcase personalized AI experiences and an x402-powered marketplace for AI creations.
Is Plastic Labs a public or private company?
Plastic Labs is a private company. It is classified as venture growth investor backed and is currently operating.
When was Plastic Labs founded?
Plastic Labs was founded in 2023. It employs 1 to 10 people.
Where is Plastic Labs based?
Plastic Labs is headquartered in New York, United States, in the North America region.
How does Plastic Labs make money?
Two revenue lines are on record. Honcho Platform (Managed SaaS) is the primary driver. The others are open Source Software.
Who are Plastic Labs's main competitors?
Direct peers on record are Zep (Zep AI), Mem0, Letta (formerly MemGPT), Cognee and Supermemory. Broad incumbents are LangChain, LlamaIndex and Pinecone. Emerging players are Inworld AI and Letta Cloud / MemGPT open source.
Does Plastic Labs have an API?
Yes. Honcho exposes multiple API endpoints: chat() for natural language queries about peers using a research agent with tool-calling capabilities; get_context() for retrieving relevant context about sessions or peers; message ingestion APIs for storing messages and deriving conclusions; and a dialectic endpoint for synthesizing answers from peer representations. The Python SDK is available as open-source. Additionally, x402 payment protocol is supported for agent-to-agent micropayments. Developer documentation is at honcho.dev.
What industry is Plastic Labs in?
Plastic Labs's product category is AI Memory Infrastructure. Its primary akta.pro industry code is HDAEANAA, End-to-End Enterprise AI Platforms (MLOps & Model Lifecycle Management), with a secondary code of HDAEANAB, Model Development & Training Platforms (AutoML, Notebooks, Feature Stores). Its NAICS code is 5132 and its SIC code is 7372.