> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.livingbrain.com/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.livingbrain.com/_mcp/server.

# Benchmarks

We measure Living Brain on LoCoMo, a public benchmark for long-term conversational memory. It gives a
system a long history of conversations, then asks questions that can only be answered by remembering the
right detail from the right session.

We report how many answers were judged correct and how much memory context the answering model read to
produce them.

* **LoCoMo accuracy**: 90.5%. Judged accuracy across all 1,540 questions
* **Less context**: 3.8×. \~6.8K tokens per question vs \~26K for full context
* **Correct answers**: 1,393. Of 1,540 questions
* **Median retrieval**: 467ms. Search plus page reads
* **Median answer**: 9.8s. Retrieval plus answer generation

## LoCoMo

[LoCoMo](https://snap-research.github.io/locomo/) contains 10 long conversations between two people,
each running across many sessions, and 1,540 questions about them. Its question types test different
kinds of remembering:

* **Single-hop**: recall one fact from one session.
* **Multi-hop**: combine facts from different sessions.
* **Temporal**: work out when something happened, or in what order.
* **Open-domain**: combine what was said with general knowledge.

**LoCoMo accuracy by question type**

|                                    | Value |
| ---------------------------------- | ----- |
| Temporal reasoning (321 questions) | 92.2% |
| Single-hop (841 questions)         | 91.7% |
| Multi-hop (282 questions)          | 91.5% |
| Open-domain (96 questions)         | 70.8% |

Batch 4fd3f746, 26 October 2026. Questions per type sum to 1,540.

Temporal, single-hop and multi-hop questions all score above 91%, and open-domain scores 70.8%.
Single-hop questions still produce the most wrong answers in absolute terms, since they are more than
half of the set.

**Memory context per question**

|                           | Value  |
| ------------------------- | ------ |
| Living Brain              | \~6.8K |
| Full conversation history | \~26K  |

Average tokens of memory placed in the answering model's context for each question.

## Why the numbers look this way

Pages are compiled when material arrives, so a question is answered from a few maintained pages
instead of every passage that mentions the topic. Related facts already sit together on one page, and
retrieval doesn't have to collect them one fragment at a time. That keeps the memory context near 6.8K
tokens, although the conversation histories average about 26K tokens.

## How we ran it

|                 | LoCoMo                                                               |
| --------------- | -------------------------------------------------------------------- |
| Dataset         | `locomo` v1.2.0, 1,540 questions                                     |
| Batch           | `4fd3f746`                                                           |
| Date            | 26 October 2026                                                      |
| Answering model | GPT-5                                                                |
| Scoring         | Each answer judged correct or incorrect against the reference answer |

Some definitions:

* **Accuracy** is correct answers divided by questions judged. Every question was judged.
* **Memory context** counts the tokens of retrieved memory given to the answering model. It doesn't
  include the instructions, the question or the model's output.
* **Full context** is the size of the full history, about 26K tokens per question. It's a size
  reference only, with no accuracy score reported for it.
* **Retrieval time** covers search and page reads. **Answer time** covers retrieval and answer
  generation together. Both are medians.

## What this benchmark doesn't cover

LoCoMo is a question-and-answer test over a fixed history. It measures whether the right evidence reaches
the answer. It doesn't measure whether a brain surfaces something at the right moment without being asked,
which is the job of the [Re-entry Priority Score](/docs/design-policies#re-entry-priority-score), or how useful its pages are to a person reading them. Each figure on this page comes from a single run.