## Performance evidence update / 18 September 2026 MODALITY PERFORMANCE / EVIDENCE REVIEWED 18 SEPTEMBER 2026 Define the useful result. Measure the complete job. A modality-specific timing is not yet linked to this guide. There may be supporting engineering measurements in the internal archive. We are reconciling their workload, result quality and comparison conditions before presenting a public chart. A benchmark for an adjacent model does not establish this model’s performance. 01 Retrieval quality 02 Parse-to-searchable time 03 Permission & update visibility Include memory, storage, energy and hardware writes ## Implementation update — 18 September 2026 Implementation snapshot / Reviewed 18 September 2026 Use stored knowledge to explain a match. Documented native search regression A documented regression indexes “automobile” and stores its lexical relationship to “car”. Exact-token search finds the first term; enabling stored synonym expansion lets the second retrieve the entity. The application uses the native Rust search API. [See the stored synonym example →](/outcomes/evidence-retrieval#example) API, release scope and evidence limits unified_search_with_store with a populated store and typed request. Example pinned to native_search_lexical_scope.rs at 7daa7908; later integration status is tracked separately. This is a scoped search fixture, not qualification of every document parser, access-control policy or assistant workflow. The reviewed search records retain open integration and broader release gates. Confirm the evaluated build and input formats. --- # Documents & text | 8DB Product reference. Reviewed 18 September 2026. Implementation and evaluation scope are stated below. Explore the models (https://8braid.com/modalities) / 8DB Documents & text Keep the source behind the answer. Preserve passages and document structure so retrieval can lead back to the material that supports a result. Explore the capabilities See a complete workload Structure · search · trace Model / Documents Conceptual diagram [Diagram: documents representation: conceptual diagram] A document is represented through its sections and passages, with a route back to the source. Capabilities What changes Input to value Alternatives Performance Adoption Related articles Reference Start with the operations your question needs. Document and hierarchical representations Native lexical retrieval Links between passages, claims and source documents Documented product scope, reviewed 17 September 2026. The underlying build and deployment must be qualified for your application. Parsing coverage depends on the input format and decomposer. Extracting a claim, generating an embedding or interpreting a document can require an external model; storage does not establish correctness of those derived claims. Connected native models Keep the next question within reach. Document retrieval can be the start of a source-backed investigation. 8DB keeps document structure and relationships available alongside vectors, records and graph dependencies. [Diagram: vectors representation: conceptual diagram] Vectors Give a similar result its context. Compare · rank · retrieve Explore model ↗ (https://8braid.com/modalities/vectors) [Diagram: graphs representation: conceptual diagram] Graphs Follow the relationship to a better answer. Connect · traverse · follow Explore model ↗ (https://8braid.com/modalities/graphs) [Diagram: tables representation: conceptual diagram] Tables Familiar rows. More context within reach. Filter · sort · project Explore model ↗ (https://8braid.com/modalities/tables) These connections show the models involved in the application design. Compare the supported combination and full workflow with a realistic alternative. Input → usable value / interactive design example Answer an engineering question with inspectable sources 0 1 Receive a report → 0 2 Keep the source → 0 3 Find passages → 0 4 Connect records → 0 5 Assemble evidence → [Diagram: documents representation: conceptual diagram] Step 1 / Documents Receive and parse a supported report A document is represented through its sections and passages, with a route back to the source. Next: Keep the source Conceptual application workflow. Qualify the supported operations, external tools, protection and persistence for the complete path. These diagrams are explanatory, not measured results. Read the complete workflow and comparison scope Receive and parse a supported report Preserve document sections and source identity Retrieve relevant passages and semantic candidates Resolve related equipment and recorded observations Assemble an answer with traceable supporting passages A useful answer includes parsing, indexing, retrieval, evidence assembly and any model call. Compare the quality and cost of that whole path; count synchronization if source documents and indexes live separately. Compare a realistic alternative. Multimodel products such as SurrealDB provide document and related data-model entry points. The choice requires comparing concrete source-preservation and query behaviour, not just counting supported labels. SurrealDB data models (https://surrealdb.com/docs/learn/data-models) Compare equivalent outputs and required guarantees, including the simplest single-product alternative where one exists. A specialist may remain the better fit when its operators, ecosystem or operational maturity are decisive. Combining native models becomes valuable when your actual question crosses them. Performance & resources Measure the work that delivers the answer. Source/section fidelity; retrieval relevance; source coverage; ingest-to-search delay; answer assembly latency; update propagation; bytes and model-call cost. This draft does not assign a new speed ratio or score to this modality. Existing measurements and source leads are retained in the evidence programme; a qualified public result must identify the particular operation and path tested. Explore the published benchmark evidence (https://8braid.com/benchmarks) What belongs inside the timer and resource budget? For input-to-value testing, declare the input state and the consumer-ready output first. Include parsing, derivation, storage/protection, index readiness, querying, result assembly and delivery as applicable. Identify amortized setup and any work deferred past the finish line. Report database-stage and full-job results separately. Do not add unrelated medians or p95 values, or multiply component speedups into a claimed workflow result. Energy needs measured joules; reduced handoffs alone do not establish a saving. Choose the path that fits your deployment. Verify the supported API/SDK, data format, persistence and indexing behaviour for the release you plan to use. Match the operation-by-modality protection profile and device qualification to the workload. Post-quantum protection (https://8braid.com/pqc) · Plan a workload evaluation (https://8braid.com/build) Related articles Take the next step into the thinking. Mechanism & application Show Your Work: Provenance as a Query, Not a Forensic Reconstruction (https://8braid.com/journal/show-your-work-provenance-as-a-query) Explore the ideas and implementation context behind this capability. Mechanism & application Give Your AI the Evidence Behind the Answer (https://8braid.com/journal/the-database-is-part-of-the-model) Explore the ideas and implementation context behind this capability. Reference / Reviewed 17 September 2026 Readable, shareable and traceable. This reference explains 8DB’s public modality description (https://8braid.com/modalities#explore) with an explicit workload and selection criteria. It is a documentary review, not fresh execution of every operator or platform. Public sources: 8DB modality scope (https://8braid.com/modalities) ; SurrealDB data models (https://surrealdb.com/docs/learn/data-models) ; the related articles above. Internal source leads and outstanding release checks remain in the editorial record. Download this reference (https://8braid.com/modality-reference/documents.md) Evaluate the question your application needs to answer. Bring the data shapes, input state, expected output and the systems you use today. Define a comparison around that complete job. Discuss your workload (https://8braid.com/build)