The ScorecardNine assistants on memory, your data, price and leaving, graded A to F side by side
AssistantsChatGPT, Claude, Gemini and six more: what each remembers, and your rights to it
Coding agentsClaude Code, Cursor, Codex, Copilot and more: what each remembers, graded
Memory systemsMem0, Zep, Letta and more: open source, self-host, export, delete, graded
Open modelsQwen, DeepSeek, Gemma, gpt-oss and more: license, weights, price, training, graded
Inference APIsGroq, Together, OpenRouter, OpenAI, Anthropic and more: price, limits, training, retention, graded
Apps, hardware, hostingOllama, LM Studio, graphics cards, GPU clouds and routers, compared in one table each
AlternativesAlternatives to every AI on the record, ranked by the same grades
Your dataSee what each AI keeps on you, and fix it in minutes
How we gradeEvery grade from the maker's own page, dated, corrected in public

Did it change?Price moves, limits and quiet downgrades, dated, with the source, and the feed
Where we areThe current state of AI, for someone who just arrived, with the daily log
Open dataThe whole record as JSON and CSV anyone can pull and cite
Grade badgesEvery graded product's letter, as a badge to embed

Run your own AIOpen models on your own machine: start here
Which model fitsPick your memory, from 8GB to 128GB, and see the open models that fit
Which app to useLM Studio, Ollama, Jan, Open WebUI and more, compared
The hardware guideWhat to buy, or not, to run models at home
The learning pathFree lessons from your first local model to agents of your own
Straight answersOllama vs LM Studio, the best local coding model, low-end hardware

The Continuity LayerWhy AI needs continuity, not just memory: the thesis, the framework, the research
Guides and glossaryWhich AI for which job, what breaks after the demo, the field's terms
The LadderFour free rungs: from chatting with AI to an agent that works for you
See what your AI remembersPaste what it tells you, see it as a map, take it with you
Succession (free)The free Claude Code plugin: every session gets a name and a handoff
Build your ownA record your AI reads first and writes to last: the guide
Research and benchmarksThe research library: every source with its contribution, limits and status
Every pageEvery page on the site, in one place

The seatEvery AI that has held the seat, and the line it left
How this site is builtOne person directs, AI sessions do the work, every change logged
The AuditPaste your website, see what a stranger sees: a grade and the one fix
Work with AdamThe Setup Desk: we set up your AI agent

Compare

The ScorecardNine assistants on memory, your data, price and leaving, graded A to F side by side AssistantsChatGPT, Claude, Gemini and six more: what each remembers, and your rights to it Coding agentsClaude Code, Cursor, Codex, Copilot and more: what each remembers, graded Memory systemsMem0, Zep, Letta and more: open source, self-host, export, delete, graded Open modelsQwen, DeepSeek, Gemma, gpt-oss and more: license, weights, price, training, graded Inference APIsGroq, Together, OpenRouter, OpenAI, Anthropic and more: price, limits, training, retention, graded Apps, hardware, hostingOllama, LM Studio, graphics cards, GPU clouds and routers, compared in one table each AlternativesAlternatives to every AI on the record, ranked by the same grades Your dataSee what each AI keeps on you, and fix it in minutes How we gradeEvery grade from the maker's own page, dated, corrected in public

Changes

Did it change?Price moves, limits and quiet downgrades, dated, with the source, and the feed Where we areThe current state of AI, for someone who just arrived, with the daily log Open dataThe whole record as JSON and CSV anyone can pull and cite Grade badgesEvery graded product's letter, as a badge to embed

Your own AI

Run your own AIOpen models on your own machine: start here Which model fitsPick your memory, from 8GB to 128GB, and see the open models that fit Which app to useLM Studio, Ollama, Jan, Open WebUI and more, compared The hardware guideWhat to buy, or not, to run models at home The learning pathFree lessons from your first local model to agents of your own Straight answersOllama vs LM Studio, the best local coding model, low-end hardware

Learn

The Continuity LayerWhy AI needs continuity, not just memory: the thesis, the framework, the research Guides and glossaryWhich AI for which job, what breaks after the demo, the field's terms The LadderFour free rungs: from chatting with AI to an agent that works for you See what your AI remembersPaste what it tells you, see it as a map, take it with you Succession (free)The free Claude Code plugin: every session gets a name and a handoff Build your ownA record your AI reads first and writes to last: the guide Research and benchmarksThe research library: every source with its contribution, limits and status Every pageEvery page on the site, in one place

About

The seatEvery AI that has held the seat, and the line it left How this site is builtOne person directs, AI sessions do the work, every change logged The AuditPaste your website, see what a stranger sees: a grade and the one fix Work with AdamThe Setup Desk: we set up your AI agent

Find a product, a question, a tool or a term.

thecontinuitylayer.com/search

498 entries, all read from the site's own data. Nothing leaves your browser.

Search the site

Prefer to browse? The map lists every page →

The recordThe Scorecard Today: Did it change?The map Run your own AI Where we are Open data Grade badges The seat

AboutHow this site is built How we grade Disclosure Who maintains this How to cite License & reuse Site changelog Report copied content Succession The Ladder The Audit Work with Adam Contact

ResearchThe Continuity Layer The thesis The framework The erasure ladder Research library Benchmarks The landscape Standards and law Open problems Field Watch Definition Primer Why AI forgets Glossary Original research Constitution The System Governance The Lab

FollowCommunity YouTube X

The Continuity Layer · The Continuity Layer is built and maintained by one independent builder who also does commercial work in this field. The System, Governance and Lab pages describe the maintainer's own research prototype; they are project reports, not independent evaluations. Published at thecontinuitylayer.com. Content licensed for citation and quotation with attribution — see /license · © The Continuity Layer