04 / AI SYSTEMS COMPARED
RiotSpark vs ChatGPT, Gemini, Copilot, Claude and Letta
Where does a self-owned AI thinking partner fit?
RiotSpark puts persistent identity, governed memory and human authority at the centre of its architecture. This comparison explains that focus alongside established assistants and stateful agent platforms, with a clear boundary between working features and future plans.
By Raziel Delacroix · RiotSpark status updated 9 September 2026 · Competitor sources reviewed 30 August 2026 · Architecture comparison, not a performance benchmark
Different products. A useful comparison.
ChatGPT, Gemini, Copilot and Claude are assistant products. Letta is a platform for stateful agents. RiotSpark is an independently developed AI core that uses a language model as its reasoning engine. Comparing them is useful when the question is how memory, identity and control work—not which name wins an intelligence score.
The established assistants already offer memory and personalisation. Letta also supports persistent identity, local deployment and model choice. RiotSpark’s focus is the combination of a protected Core, reviewable knowledge, approval rules and recorded reasons for remembering.
Scope: the named assistant products and documented Letta capabilities, not every provider API, enterprise deployment or third-party extension. Features vary by plan, region, account and configuration. Descriptions below link to official documentation; absence of a feature from this page does not establish that a product lacks it.
AI memory, identity and ownership at a glance
Read each product in its own context. On smaller screens, scroll the table horizontally.
RiotSpark and five related AI products — documented capabilities and development status| System / purpose | Memory and continuity | Ownership and scope | Evidence |
|---|
| RiotSparkIndependent AI core in development | External identity Core; persistent conversations; governed capture, tiered memory, bounded retrieval, Knowledge Studio, correction and auditable revision history. | Local Ollama runtime verified. Broader engine compatibility needs adapter integration and validation. | Project milestones |
|---|
| ChatGPTHosted AI assistant | Memory, custom instructions, review controls and explanations of memory sources. | Personalisation is managed within the ChatGPT product. | OpenAI documentation |
|---|
| GeminiGoogle’s AI assistant | Past-chat personalisation and response instructions for eligible accounts. | Availability depends on account and feature settings. | Google documentation |
|---|
| Microsoft CopilotMicrosoft’s consumer AI assistant | Personalisation can remember details; users can edit, delete or disable memory. | This comparison covers consumer Copilot, not GitHub Copilot or the full Microsoft 365 product range. | Microsoft documentation |
|---|
| ClaudeAnthropic’s AI assistant | Cross-chat memory, project context and review controls; memory import and export are also documented. | Transfer of memory text is distinct from running an independent identity and governance runtime. | Memory · Transfer |
|---|
| LettaStateful agent platform | Persistent, versioned memory through MemFS, including identity files and on-demand reference material. | Supports local or self-hosted agents and multiple model providers. | Memory · Hosting · Models |
|---|
What RiotSpark does today—and what comes next
IMPLEMENTED · THROUGH P2.6A working governed foundation
Local chat through Ollama, an external identity and personality Core, conversations that resume across restarts, and Knowledge Studio for manual memory capture, candidate review, classification and approval.
P2.5 added tier inspection and placement, bounded tier-aware retrieval, owner-authored distillation with source links, and storage and growth visibility. P2.6 adds scheduled reviews, revision comparison, governed editing, correction and supersession, subject inspection, conflict visibility and preserved audit history.
P2.6 is complete and owner-accepted. Checkpoint validation covered 330 automated tests: 328 passed and 2 were skipped, with live SQLite integrity verified.
View completed milestones →IN DEVELOPMENT · P2.7Make identity inspectable and continuous
Personality Continuity and Behaviour Studio is the active checkpoint. Its scope includes interaction preferences, approved adaptive traits, personality and relationship-matrix inspection, proposed-change evidence, response previews, owner approval, version history and rollback.
P2.7 is in progress. Its acceptance goal is recognisable continuity across restarts and approved model changes, with every behavioural change reversible and attributable and no Core drift through ordinary editing.
Follow the current roadmap →LONGER-TERM DIRECTIONContinuity across richer interfaces
The aim is to preserve RiotSpark’s identity as reasoning engines and interfaces evolve. Voice, visual presence, embodied interfaces and robotics remain future directions.
No claim of universal model compatibility, autonomous company operation or superiority over other assistants is made here. Model changes still need compatibility and behaviour checks.
Read the governing principles →
How RiotSpark compares with each system
RiotSpark vs ChatGPT
ChatGPT provides memory and custom instructions within a hosted assistant, including controls for reviewing remembered context and explanations of memory sources. RiotSpark concentrates on keeping its identity and approval workflow in an owner-operated core outside the model. The distinction is the architecture and governance boundary, not whether ChatGPT can remember. [1]
RiotSpark vs Gemini
Gemini Apps can personalise responses using past chats and explicit instructions, subject to eligibility and settings. RiotSpark’s current implementation focuses on a local runtime and explicitly governed knowledge capture. Someone choosing between them should consider whether they want an existing assistant service or to follow the development of an independently controlled AI core. [2]
RiotSpark vs Microsoft Copilot
Consumer Copilot can remember useful personal details and lets users change or disable that memory. RiotSpark instead makes its protected Core and knowledge approval process part of its own runtime. Copilot’s consumer, Microsoft 365 and GitHub products have different purposes and controls; this page does not treat them as interchangeable. [3]
RiotSpark vs Claude
Claude documents cross-chat memory, project-specific context and memory controls, as well as importing and exporting memory. That means portability is not an all-or-nothing distinction. RiotSpark’s aim is to preserve the governing system itself across engine changes, rather than only move remembered text between services. Wider model support remains something to implement and verify. [4] [5]
RiotSpark vs Letta
Letta is the closest architectural comparison here. Its MemFS stores versioned memory, with identity files in active context and deeper material read when needed. It also supports local or self-hosted agents and multiple model providers. These are real areas of overlap, not exclusive RiotSpark features. [6] [7] [8]
RiotSpark’s emphasis is its particular governance policy: protected identity, owner approval, knowledge classification and a mandatory why-layer. Whether that policy serves a use case better requires practical evaluation. Letta can be extended; this page does not claim it cannot implement similar controls.
Why follow RiotSpark?
If you care about a persistent AI companion whose identity, memory rules and development direction remain under human control, RiotSpark is a project to watch. Its public value today is a working local foundation and a visible development path. It is not presented as a finished replacement for the established products above.
No comparative accuracy, speed, safety or cost benchmark has been published here. The project’s own test counts validate its stated release checks; they do not measure performance against ChatGPT, Gemini, Copilot, Claude or Letta. Local operation also brings responsibility for hardware, updates and backups.