FlowiseAI/Flowise · error · Error

Invalid Similarity Metric should be one of 'cosine' | 'eucli

Error message

Invalid Similarity Metric should be one of 'cosine' | 'euclidean' | 'dot_product'

What it means

Thrown in Astra vectorStoreMethods.upsert when `similarityMetric` is set but not one of 'cosine', 'euclidean', 'dot_product'. The validation runs before any Astra client is created, so it is purely an input-config check. The same check exists in init (error 498).

Source

Thrown at packages/components/nodes/vectorstores/Astra/Astra.ts:109

                name: 'vectorStore',
                baseClasses: [this.type, ...getBaseClasses(AstraDBVectorStore)]
            }
        ]
    }

    //@ts-ignore
    vectorStoreMethods = {
        async upsert(nodeData: INodeData, options: ICommonObject): Promise<Partial<IndexingResult>> {
            const docs = nodeData.inputs?.document as Document[]
            const embeddings = nodeData.inputs?.embeddings as Embeddings
            const vectorDimension = nodeData.inputs?.vectorDimension as number
            const astraCollection = nodeData.inputs?.astraCollection as string
            const similarityMetric = nodeData.inputs?.similarityMetric as 'cosine' | 'euclidean' | 'dot_product' | undefined
            const credentialData = await getCredentialData(nodeData.credential ?? '', options)

            const expectedSimilarityMetric = ['cosine', 'euclidean', 'dot_product']
            if (similarityMetric && !expectedSimilarityMetric.includes(similarityMetric)) {
                throw new Error(`Invalid Similarity Metric should be one of 'cosine' | 'euclidean' | 'dot_product'`)
            }

            const clientConfig = {
                token: credentialData?.applicationToken,
                endpoint: credentialData?.dbEndPoint
            }

            const astraConfig: AstraLibArgs = {
                ...clientConfig,
                collection: astraCollection ?? credentialData.collectionName ?? 'flowise_test',
                collectionOptions: {
                    vector: {
                        dimension: vectorDimension ?? 1536,
                        metric: similarityMetric ?? 'cosine'
                    }
                }
            }

View on GitHub (pinned to abe4a8601a)

Solutions

1. Set the similarity metric to one of 'cosine', 'euclidean', or 'dot_product'.
2. Leave the field unset to use the Astra default rather than supplying an invalid value.
3. Normalize input to lowercase before validation if casing is the issue (but the allowed set is lowercase only).

Example fix

// before
similarityMetric = 'manhattan' // throws: Invalid Similarity Metric ...

// after
similarityMetric = 'cosine' // or 'euclidean' | 'dot_product'

const ALLOWED_ASTRA_METRICS = ['cosine', 'euclidean', 'dot_product'] as const
type AstraMetric = typeof ALLOWED_ASTRA_METRICS[number]
function normalizeMetric(v: unknown): AstraMetric | undefined {
  if (v == null || v === '') return undefined
  const lower = String(v).toLowerCase()
  if (!ALLOWED_ASTRA_METRICS.includes(lower as AstraMetric)) {
    throw new Error(`similarityMetric must be one of ${ALLOWED_ASTRA_METRICS.join('|')}, got: ${String(v)}`)
  }
  return lower as AstraMetric
}
// before upsert
normalizeMetric(nodeData.inputs?.similarityMetric)

const isAstraMetric = (v: unknown): v is 'cosine' | 'euclidean' | 'dot_product' =>
  typeof v === 'string' && ['cosine', 'euclidean', 'dot_product'].includes(v)

try { return await astraNode.vectorStoreMethods.upsert(nodeData, options) }
catch (e) {
  if (/Invalid Similarity Metric/.test((e as Error).message)) {
    throw new Error('Set similarityMetric to cosine|euclidean|dot_product (or leave unset)')
  }
  throw e
}

When it happens

Trigger: A user selects or types a similarity metric value outside the allowed set (e.g. 'manhattan', 'Cosine' with wrong casing, 'dot', empty-but-defined).

Common situations: Custom metric typed manually; case mismatch; stale flow referencing a previously-supported metric; typo in API-driven flow creation.

Related errors

• Invalid path: path must be within allowed directories (${all
• Model is required
• Model is required
• Scenarios are required
• Invalid JSON in executeFlowOverrideConfig: ${parseError.mess

AI-assisted analysis of FlowiseAI/Flowise@abe4a8601a (2026-08-12). Data as JSON: /api/errors/4217040e000e2e80. Report an issue: GitHub.