CLI command reference

pc login

# Log in via browser
pc login

# Then set target context
pc target -o "my-org" -p "my-project"

pc logout

# Clear all credentials and context
pc logout

pc target [flags]

# Interactive targeting after login
pc login
pc target

# Set specific organization and project
pc target -o "my-org" -p "my-project"

# Show current context
pc target --show

# Clear all context
pc target --clear

pc version

# Display version information
pc version

pc whoami

pc whoami

pc auth clear [flags]

# Clear only the default (manually specified) API key
pc auth clear --api-key
pc auth status

# Clear service account
pc auth clear --service-account

pc auth configure [flags]

# Service account setup (auto-targets org and prompts for project)
pc auth configure --client-id my-id --client-secret my-secret

# Service account with specific project
pc auth configure \
  --client-id my-id \
  --client-secret my-secret \
  -p proj-123

# Default API key (overrides any target context)
pc auth configure --api-key pcsk_abc123

pc auth local-keys list [flags]

# List all locally managed keys
pc auth local-keys list

# Show key values
pc auth local-keys list --reveal

# After storing a key
pc api-key create -n "my-key" --store
pc auth local-keys list

pc auth local-keys prune [flags]

# Preview deletions
pc auth local-keys prune --dry-run

# Delete CLI-created keys only
pc auth local-keys prune -o cli --skip-confirmation

# Delete for specific project
pc auth local-keys prune --id proj-abc123

# Before/after check
pc auth local-keys list
pc auth local-keys prune -o cli
pc auth local-keys list

pc auth login
pc login  # shorthand

# Login and set target
pc auth login
pc target -o "my-org" -p "my-project"
pc index list

pc auth logout

# Clear all credentials and context
pc auth logout

pc auth status [flags]

# Check status after login
pc auth login
pc auth status

# JSON output for scripting
pc auth status --json

pc auth whoami

pc auth whoami

pc index configure [flags]

# Enable deletion protection
pc index configure --index-name my-index --deletion-protection enabled

# Add tags
pc index configure --index-name my-index --tags environment=production,team=ml

# Switch to dedicated read capacity
pc index configure --index-name my-index \
  --read-mode dedicated \
  --read-node-type b1 \
  --read-shards 2 \
  --read-replicas 2

# Verify changes
pc index describe --index-name my-index

pc index create [flags]

# Create serverless index
pc index create -n my-index -d 1536 -m cosine -c aws -r us-east-1

# Create sparse vector index
pc index create -n sparse-index -m dotproduct -c aws -r us-east-1 --vector-type sparse

# With integrated embedding model
pc index create \
  -n my-index \
  -m cosine \
  -c aws \
  -r us-east-1 \
  --model multilingual-e5-large \
  --field-map text=chunk_text

# With dedicated read capacity
pc index create \
  -n my-index \
  -d 1536 \
  -m cosine \
  -c aws \
  -r us-east-1 \
  --read-mode dedicated \
  --read-node-type b1 \
  --read-shards 2 \
  --read-replicas 2

# With deletion protection
pc index create \
  -n my-index \
  -d 1536 \
  -m cosine \
  -c aws \
  -r us-west-2 \
  --deletion-protection enabled

# From collection
pc index create \
  -n my-index \
  -d 1536 \
  -m cosine \
  -c aws \
  -r eu-west-1 \
  --source-collection my-collection

# Create pod-based index
pc index create \
  -n my-index \
  -d 1536 \
  -m cosine \
  --environment us-east-1-aws \
  --pod-type p1.x1 \
  --shards 1 \
  --replicas 1

pc index delete [flags]

# Delete an index (prompts for confirmation)
pc index delete --index-name my-index

# Skip the confirmation prompt
pc index delete --index-name my-index --skip-confirmation

# List before and after
pc index list
pc index delete --index-name test-index --skip-confirmation
pc index list

pc index describe [flags]

# Describe an index
pc index describe -i my-index

# JSON output
pc index describe -i my-index -j

# Check newly created index
pc index create -n test-index -d 1536 -m cosine -c aws -r us-east-1
pc index describe -i test-index

pc index stats [flags]

# Get stats for an index
pc index stats -i my-index

# Get stats with a metadata filter
pc index stats -i my-index --filter '{"genre":{"$eq":"rock"}}'

# Filter from file
pc index stats -i my-index --filter ./filter.json

# JSON output
pc index stats -i my-index -j

pc index list [flags]

# List all indexes
pc index list

# Show additional details
pc index list --wide

# JSON output for scripting
pc index list -j

# After creating indexes
pc index create -n test-1 -d 768 -m cosine -c aws -r us-east-1
pc index list

pc index namespace create [flags]

# Create a namespace
pc index namespace create -i my-index --name tenant-a

# Create with metadata schema (comma-separated list of filterable metadata fields)
pc index namespace create -i my-index --name tenant-b --schema "category,brand"

# JSON output
pc index namespace create -i my-index --name tenant-c -j

pc index namespace delete [flags]

# Delete a namespace (prompts for confirmation)
pc index namespace delete -i my-index --name tenant-a

# Delete the default namespace
pc index namespace delete -i my-index --name __default__

# Skip the confirmation prompt
pc index namespace delete -i my-index --name tenant-a --skip-confirmation

pc index namespace describe [flags]

# Describe a namespace
pc index namespace describe -i my-index --name tenant-a

# Describe the default namespace
pc index namespace describe -i my-index --name __default__

# JSON output
pc index namespace describe -i my-index --name tenant-a -j

pc index namespace list [flags]

# List all namespaces
pc index namespace list -i my-index

# Filter by prefix
pc index namespace list -i my-index --prefix "tenant-"

# Limit results
pc index namespace list -i my-index --limit 10

# JSON output
pc index namespace list -i my-index -j

pc index vector delete [flags]

# Delete specific vectors
pc index vector delete -i my-index --ids '["id1"]'

# Delete multiple vectors (inline JSON array, or JSON array in a file)
pc index vector delete -i my-index --ids '["id1", "id2"]'

# Delete by filter
pc index vector delete -i my-index --filter '{"genre":"classical"}'

# Delete all vectors in a namespace (prompts for confirmation)
pc index vector delete -i my-index --namespace old-data --all-vectors

# Delete all vectors without the prompt
pc index vector delete -i my-index --namespace old-data --all-vectors --skip-confirmation

pc index vector fetch [flags]

# Fetch specific vectors by ID
pc index vector fetch -i my-index --ids '["123","456","789"]'

# Fetch from a file
pc index vector fetch -i my-index --ids ./ids.json

# Fetch by metadata filter
pc index vector fetch -i my-index --filter '{"genre":{"$eq":"rock"}}'

# Fetch from a namespace
pc index vector fetch -i my-index --namespace tenant-a --ids '["doc-123"]'

# JSON output
pc index vector fetch -i my-index --ids '["vec1"]' -j

pc index vector list [flags]

# List vector IDs
pc index vector list -i my-index

# List from a namespace with limit
pc index vector list -i my-index --namespace tenant-a --limit 50

# JSON output
pc index vector list -i my-index -j

pc index vector query [flags]

# Query by vector ID
pc index vector query -i my-index --id "doc-123" -k 10 --include-metadata

# Query by vector values
pc index vector query -i my-index --vector '[0.1, 0.2, 0.3]' -k 25

# Query with metadata filter
pc index vector query -i my-index \
  --vector '[0.1, 0.2, 0.3]' \
  --filter '{"genre":{"$eq":"sci-fi"}}' \
  --include-metadata

# Query from file (file contains a JSON array that specifies the query vector)
pc index vector query -i my-index --vector ./embedding.json -k 20

# Query with sparse vectors (inline)
pc index vector query -i my-index \
  --sparse-indices '[0, 5, 12]' \
  --sparse-values '[0.5, 0.3, 0.8]' \
  -k 15

# Query with sparse vectors from files
# indices.json: [0, 5, 12]
# values.json: [0.5, 0.3, 0.8]
pc index vector query -i my-index \
  --sparse-indices ./indices.json \
  --sparse-values ./values.json \
  -k 15

# Query from stdin (extract embedding from a document)
# doc.json: {"id": "doc-123", "embedding": [0.1, 0.2, 0.3], "text": "..."}
jq -c '.embedding' doc.json | pc index vector query -i my-index --vector - -k 10

pc index vector update [flags]

# Update metadata for a single vector
pc index vector update -i my-index --id "vec1" --metadata '{"category":"updated"}'

# Update values for a single vector
pc index vector update -i my-index --id "vec1" --values '[0.2, 0.3, 0.4]'

# Update sparse values
# indices.json: [0, 5, 12]
# values.json: [0.5, 0.3, 0.8]
pc index vector update -i my-index --id "vec1" \
  --sparse-indices ./indices.json \
  --sparse-values ./values.json

# Bulk update metadata by filter (preview first)
pc index vector update -i my-index \
  --filter '{"genre":{"$eq":"sci-fi"}}' \
  --metadata '{"genre":"fantasy"}' \
  --dry-run

# Apply the bulk update
pc index vector update -i my-index \
  --filter '{"genre":{"$eq":"sci-fi"}}' \
  --metadata '{"genre":"fantasy"}'

pc index vector upsert [flags]

# Upsert from JSON file (with "vectors" array)
# vectors.json: {"vectors": [{"id": "vec1", "values": [0.1, 0.2, 0.3], "metadata": {"genre": "comedy"}}]}
pc index vector upsert -i my-index --body ./vectors.json

# Upsert with inline JSON
pc index vector upsert -i my-index --body '{"vectors": [{"id": "vec1", "values": [0.1, 0.2, 0.3], "metadata": {"genre": "comedy"}}]}'

# Upsert from JSONL file (one vector per line)
# vectors.jsonl: {"id": "vec1", "values": [0.1, 0.2, 0.3]}
#                {"id": "vec2", "values": [0.4, 0.5, 0.6]}
pc index vector upsert -i my-index --body ./vectors.jsonl

# Upsert from stdin (same format as JSON or JSONL file)
cat vectors.json | pc index vector upsert -i my-index --body -

# Custom batch size (default: 500, max: 1000 per API request)
pc index vector upsert -i my-index --body ./vectors.json --batch-size 1000

pc index record search [flags]

# Text search (integrated embedding indexes only)
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --top-k 10
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --namespace my-namespace --top-k 5
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --filter '{"category":"health"}' --top-k 10
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --fields '["_id","chunk_text","category"]' --top-k 10
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --rerank '{"model":"bge-reranker-v2-m3","rank_fields":["chunk_text"]}' --top-k 10
pc index record search --index-name my-index --inputs '{"text":"disease prevention"}' --match-terms '{"terms":["vaccine","prevention"]}' --top-k 10

# Search by record ID for query vector (all index types)
pc index record search --index-name my-index --id rec-123 --top-k 10

# Vector search (all index types)
pc index record search --index-name my-index --vector '[0.1,0.2,0.3]' --top-k 10
pc index record search --index-name my-index --sparse-indices '[1,5,9]' --sparse-values '[0.4,0.8,0.2]' --top-k 10
pc index record search --index-name my-index --vector '[0.1,0.2,0.3]' --sparse-indices '[1,5,9]' --sparse-values '[0.4,0.8,0.2]' --top-k 10

# Full request body (must include top_k in the body)
pc index record search --index-name my-index --body ./search.json

pc index record upsert [flags]

# Upsert from a JSON file
pc index record upsert --index-name my-index --body ./records.json

# Upsert from a JSONL file into a specific namespace
pc index record upsert --index-name my-index --namespace my-namespace --body ./records.jsonl

# Upsert from stdin
cat records.jsonl | pc index record upsert --index-name my-index --namespace my-namespace --body -

pc index collection create [flags]

# Create a collection from a pod-based index
pc index collection create --name my-collection --source my-index

pc index collection delete [flags]

# Delete a collection (prompts for confirmation)
pc index collection delete --name my-collection

# Skip the confirmation prompt
pc index collection delete --name my-collection --skip-confirmation

pc index collection describe [flags]

# Describe a collection
pc index collection describe --name my-collection

pc index collection list [flags]

# List all collections
pc index collection list

pc index backup create [flags]

# Create a backup
pc index backup create -i my-index

# Create with name and description
pc index backup create -i my-index --name "nightly-backup" -d "Nightly backup before deployment"

# JSON output
pc index backup create -i my-index -j

pc index backup delete [flags]

# Delete a backup by ID (prompts for confirmation)
pc index backup delete -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f

# Skip the confirmation prompt
pc index backup delete -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f --skip-confirmation

pc index backup describe [flags]

# Describe a backup
pc index backup describe -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f

# JSON output
pc index backup describe -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f -j

pc index backup list [flags]

# List all backups in the project
pc index backup list

# List backups for a specific index
pc index backup list --index-name my-index

# Limit results
pc index backup list --limit 10

# JSON output
pc index backup list -j

pc index restore [flags]

# Restore an index from a backup
pc index restore -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f -n restored-index

# Restore with tags and deletion protection
pc index restore -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f -n restored-index \
  --tags env=prod,team=search \
  --deletion-protection enabled

# JSON output
pc index restore -i c84725e5-5956-41ba-ab62-21ac7b5f2a2f -n restored-index -j

pc index restore describe [flags]

# Describe a restore job
pc index restore describe -i rj-abc123

# JSON output
pc index restore describe -i rj-abc123 -j

pc index restore list [flags]

# List restore jobs
pc index restore list

# Limit results
pc index restore list --limit 10

# JSON output
pc index restore list -j