Cogni — MCP memory server for Claude, ChatGPT and AI agents<br>C" />

The cognition layer for LLMs

More than memory.<br>Memory that reasons.

Cogni is an MCP server that gives any LLM a connected memory that grows as you work. Its<br>entity-graph spreading activation follows the chain of facts a<br>plain vector store can't reach, so your agent reasons over what it remembers instead of just looking<br>things up.

0.85

Cogni recall, reassembling a chain<br>scattered across separate documents

vs

0.65

plain vector search,<br>same store, same eight rows

Chain-recall@8: the share of a four-document chain a single recall call surfaces — Cogni finds<br>3.4 of the 4, a vector store finds 2.6. Not answer accuracy.<br>Deterministic, no model in the loop, on the shipped recall path at the settings every account gets:<br>5 chains &times; 3 seeds against 300 distractor documents. Both sides<br>return the same eight rows, so none of the margin is bought with extra context.<br>Every number, and where we tie or come second.

Get started free<br>Connect in 60 seconds

✓ Works with any model: Claude, GPT, Gemini, local<br>✓ No LLM and no GPU in the retrieval path<br>✓ No vector DB to run<br>✓ No API key to paste

A vector store retrieves. Cogni connects.

The difference is one clean, measured capability: entity-graph spreading activation for cross-vocabulary retrieval — questions whose answer shares no words with the question, where a single similarity lookup scores zero no matter how wide you set k.

Accuracy · one recall call

Ask something whose answer lives in a document sharing no vocabulary with the question — a part number that leads to a batch that leads to a name. In a single retrieval call Cogni follows the entity chain. A similarity search matches the first link and has nothing to take it to the second, so on one call it scores zero.

Cogni

0.48

Vector RAG

0.00

Measured end to end over MCP with GPT-4o answering, one recall call, on Cogni's default mode at the settings every account gets — mean of 14 runs, 45 questions each. The zero is the structural part: turning Cogni's own graph off reproduces it exactly, on the same store with the same embeddings and the same prompt, so the difference is the traversal and nothing else.

Cost · full agentic loop

Let the model keep calling instead of stopping at one, and a vector store recovers to near-parity — it eventually stumbles onto each link. That is the honest shape of the result, and it is still the argument: the answer is reachable either way; what differs is how many round trips you pay for it.

Cogni returns the chain on the first call. The alternative gets there by iterating, and every iteration is another request, another context window, and another bill from whichever model you are using.

The gap widens with the length of the chain: each extra hop is one more round trip a similarity search has to discover, and one Cogni already walked.

A full vector search runs underneath the graph. Cogni's default recall runs a dense-vector search and the spreading graph, then returns both — so adding Cogni does not take away the retrieval you already had. It is a floor you can check in one call, not a promise: on needle-in-a-haystack recall the two are level at 1.00 to 1.00, at every haystack length we measured.

The big one

Works with any model. No LLM in the loop.

No language model and no GPU anywhere in the retrieval path.

Nothing rewrites, summarises, or judges your memories on the way in, and nothing is generated on the<br>way out. Recall is a deterministic graph traversal : the same<br>question over the same store returns the same rows, every time, with no second model's opinion in<br>between. That is why the advantage travels — it lives in the retrieval, not in the model. And<br>it is why a small model gains the most: clean connected context was the piece it was missing.

0.16 &rarr; 0.53

the same local 8B on multi-hop deduction: the raw corpus in its context, then the eight rows Cogni returns

0.00

what a plain vector search scores on the same questions with the same model — while holding nearly twice the rows

18&times; less

context than a frontier model reading the whole corpus: 239 tokens a query against 4,238

To be straight about the ceiling: GPT-4o with the entire corpus in context scores 0.99, so this is not parity — it is a small local model doing work it could not do at all before, on a fraction of the context. Feed that same 8B the exact chain by hand and it scores 0.97, which says the remaining gap is retrieval rather than reasoning.

Multi-hop deductive benchmark, 12 chains &times; 3 seeds against 250 distractor documents, llama3.1:8b answering, measured through the shipped retrieval path at the settings every account gets — the results file records no deviation from them. The vector baseline was given 15 rows to Cogni's 8. Because Cogni's tools are called by your model, real-world consistency depends on the model and how it's prompted. Why results may vary.

An agent that stops forgetting

Recall...