spring-projects/spring-ai · error · IllegalArgumentException

Invalid value type for EQ. Can either be a string or Number

Error message

Invalid value type for EQ. Can either be a string or Number

What it means

buildEqCondition translates an EQ comparison into a Qdrant match condition, accepting only String or Number values; anything else (Boolean, null, Date, collection) throws this IllegalArgumentException. Qdrant's match() field conditions only support those scalar types directly.

Source

Thrown at vector-stores/spring-ai-qdrant-store/src/main/java/org/springframework/ai/vectorstore/qdrant/QdrantFilterExpressionConverter.java:102

			case LT -> buildLtCondition(key, value);
			case LTE -> buildLteCondition(key, value);
			case IN -> buildInCondition(key, value);
			case NIN -> buildNInCondition(key, value);
			default -> throw new RuntimeException("Unsupported expression type: " + type);
		};
	}

	protected Condition buildEqCondition(Key key, Value value) {
		String identifier = doKey(key);
		if (value.value() instanceof String valueStr) {
			return io.qdrant.client.ConditionFactory.matchKeyword(identifier, valueStr);
		}
		else if (value.value() instanceof Number valueNum) {
			long lValue = Long.parseLong(valueNum.toString());
			return io.qdrant.client.ConditionFactory.match(identifier, lValue);
		}

		throw new IllegalArgumentException("Invalid value type for EQ. Can either be a string or Number");

	}

	protected Condition buildNeCondition(Key key, Value value) {
		String identifier = doKey(key);
		if (value.value() instanceof String valueStr) {
			return io.qdrant.client.ConditionFactory.filter(Filter.newBuilder()
				.addMustNot(io.qdrant.client.ConditionFactory.matchKeyword(identifier, valueStr))
				.build());
		}
		else if (value.value() instanceof Number valueNum) {
			long lValue = Long.parseLong(valueNum.toString());
			Condition condition = io.qdrant.client.ConditionFactory.match(identifier, lValue);
			return io.qdrant.client.ConditionFactory.filter(Filter.newBuilder().addMustNot(condition).build());
		}

		throw new IllegalArgumentException("Invalid value type for NEQ. Can either be a string or Number");

View on GitHub (pinned to 98a7beda4f)

Solutions

  1. Replace boolean equality with a string/int: eq("active", "true") or eq("active", 1), and store metadata accordingly.
  2. Convert dates to epoch-millis Numbers before filtering.
  3. For list membership use IN instead of EQ.
  4. Normalize document metadata on ingestion so filtered fields are always String or Number.

Example fix

// before
Filter.builder().eq("isActive", true).build();
// after
Filter.builder().eq("isActive", "true").build();
Defensive patterns

Strategy: type-guard

Validate before calling

static void checkEqOperand(Object v) {
    if (!(v instanceof String) && !(v instanceof Number)) {
        throw new IllegalArgumentException("EQ operand must be String or Number, got " + (v == null ? "null" : v.getClass()));
    }
}

Type guard

static boolean isEqCompatible(Object v) {
    return v instanceof String || v instanceof Number;
}

Try / catch

try {
    vectorStore.similaritySearch(req);
} catch (IllegalArgumentException ex) {
    if (ex.getMessage() != null && ex.getMessage().startsWith("Invalid value type for EQ")) {
        throw new IllegalArgumentException("Metadata field must be String or Number for EQ", ex);
    }
    throw ex;
}

Prevention

When it happens

Trigger: filterExpression(EQ) where the right-hand Value wraps a Boolean (e.g. eq("isActive", true)), null, a date object, or a List — e.g. Filter.builder().eq("active", true).build() passed to the Qdrant store.

Common situations: Document metadata contains booleans or dates; developers write eq("flag", true) expecting Qdrant to handle it. Also common when metadata values are auto-extracted from JSON where a field is an object/array rather than a scalar.

Understand the failure class

Background: Type mismatch errors: IllegalArgumentException, TypeError and type guards across 150 open-source libraries — this error's family across 150 libraries.

Related errors


AI-assisted analysis of spring-projects/spring-ai@98a7beda4f (2026-09-11). Data as JSON: /api/errors/fdb237b7901b5541. Report an issue: GitHub.