Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.qwedai.com/llms.txt

Use this file to discover all available pages before exploring further.

Engine overview

EngineTechnologyKey features
MathSymPy + DecimalCalculus, matrix ops, NPV/IRR, statistics
LogicZ3 theorem proverForAll/Exists quantifiers, BitVectors, arrays
ReasoningMulti-LLM + cacheChain-of-thought validation, result caching
StatsWasm/Docker sandboxSecure code execution with AST validation
FactTF-IDF + NLPSemantic similarity, entity matching, citations
GraphKnowledge graphTriple verification (subject-predicate-object)
CodeMulti-lang ASTPython, JavaScript, Java, Go security analysis
SQLSQLGlot ASTComplexity limits, cost estimation, schema validation
TaintData flow analysisTrace untrusted inputs to sensitive sinks (e.g., exec, os.system, SQL queries)
ImageDeterministic + VLMMetadata extraction, size verification, multi-VLM
SchemaPydantic + MathJSON structure + embedded calculation checks

Deterministic-first philosophy

All engines follow a deterministic-first approach:
  1. Try deterministic methods first (100% reproducible)
  2. Fall back to LLM only when necessary
  3. Discount LLM confidence when used
# Example: fact verification is deterministic
result = client.verify_fact(
    claim="Paris is in France",
    context="Paris is the capital of France."
)
# Uses TF-IDF similarity + entity matching
# No LLM needed for most claims
See Determinism guarantee for the full engine classification and how to check if a response is SYMBOLIC or HEURISTIC.

Engine selection

QWED auto-detects the appropriate engine:
Content patternDetected engine
2+2=4, sqrt(16), derivativeMath
(AND ...), ForAll, ExistsLogic
SELECT, INSERT, DROPSQL
````python, import, function`Code
Claims with contextFact
Image bytes + claimImage
Or specify explicitly:
result = client.verify(query, type="math")

Engine documentation

Deterministic engines

Security engines

Data verification engines

Orchestration engines