What is Pictaria Server?

Pictaria Server is a self-hosted web application that connects to your Immich photo library to provide deep insights, AI-powered enrichment, and curation tools. It runs as a single Node.js process (Docker or bare metal) with zero npm dependencies. Key capabilities include: Insights for collection statistics and timeline exploration; Enrich for AI photo classification using local models (Ollama, LM Studio, llama.cpp) or cloud providers (OpenAI, OpenRouter, Venice); Curate for human review of photo stacks with AI-assisted keeper suggestions; Smart Albums that keep real Immich albums synchronized to saved search rules; Frame Remote for live control of Pictaria Frame devices; Voice features for photo Q&A and text-to-speech; and Activity logging for audit trails. All configuration is managed through a web UI with runtime changes, no restart required. The server uses SQLite for local state and requires an Immich 2.0+ instance. Licensed under AGPL-3.0-only.

Should you self-host Pictaria Server?

Pictaria Server scores 48/100 on selfhost.directory (how scoring works) and ranks #25 of 41 in Photo Management. It has 196 GitHub stars.

It shipped 5 stable releases in the past year. The latest, v1.2.1, came out on 12 September 2026. New versions land roughly every 3 days.

Setup is moderate: expect some configuration beyond a single command. An official Docker image is available. It is AGPL-3.0-licensed and written mostly in JavaScript.

Its closest alternative on selfhost.directory is Ente (81/100, 29 693 stars). It scores higher overall, so compare the two before you pick. See Pictaria Server vs Ente side by side.

Features

Activity Log
AI Enrichment
Ambient Display
Automatic Backups
Frame Metrics
Frame Remote
Human Curation
Insights
Runtime Settings
Smart Albums
Voice assistant

How to install Pictaria Server

Install Pictaria Server with Docker Compose

Docker and Docker Compose installed

Environment variables

IMMICH_BASE_URL, IMMICH_API_KEY, APP_PASSWORD, ALLOW_INSECURE_OPEN, SESSION_COOKIE_SECURE, BROWSER_ALLOWED_HOSTS, TRUSTED_PROXY_IPS, BACKUP_DIR, BACKUP_ENABLED, BACKUP_INTERVAL_HOURS, BACKUP_KEEP, ENRICH_ENABLED, ENRICH_SCHEDULE_ENABLED, ENRICH_SCHEDULE_TIME, ENRICH_SCHEDULE_TIME_ZONE, ENRICH_SCHEDULE_PHOTO_BUDGET, ENRICH_HISTORY_RUNS, ENRICH_HISTORY_LOGS, CAPTION_WRITEBACK, ENRICH_EMBEDDINGS_ENABLED, IMMICH_ML_URL, ENRICH_EMBEDDINGS_MODEL, IMAGE_SOURCE, INFERENCE_HOST_LABEL, MAX_FAILURES_PER_ASSET, PROMPT_VERSION, CURATE_BURST_GROUPING, CURATE_EMBEDDING_STACKS, REFEREE_GROUP_BUDGET_MB, OPENAI_API_KEY, DEFAULT_PROVIDER, CURATE_REFEREE_ENABLED, CURATE_STACK_REFEREE_ENABLED, CURATE_STACK_REFEREE_SCOPE, CURATE_KEEPER_REFEREE_ENABLED, CURATE_REFEREE_PROVIDER, CURATE_REFEREE_MODEL, TTS_PROVIDER, STT_PROVIDER, OPENAI_TTS_MODEL, OPENAI_TTS_VOICE, OPENAI_TTS_SPEED, OPENAI_TTS_FORMAT, OPENAI_TTS_INSTRUCTIONS, ELEVENLABS_API_KEY, ELEVENLABS_VOICE_ID, ELEVENLABS_TTS_MODEL, ELEVENLABS_OUTPUT_FORMAT, ELEVENLABS_REQUEST_TIMEOUT_MS, VOICE_PROSE_PROVIDER, VOICE_INTERESTING_MODEL, VOICE_ASK_MODEL, VOICE_PROSE_TIMEOUT_MS, OPENAI_INTERESTING_MODEL, OPENAI_INTERESTING_IMAGE_DETAIL, OPENAI_INTERESTING_MAX_OUTPUT_TOKENS, OPENAI_ASK_MODEL, OPENAI_ASK_MAX_OUTPUT_TOKENS, OPENAI_REQUEST_TIMEOUT_MS, VENICE_API_KEY, VENICE_MODEL, VENICE_BASE_URL, OPENROUTER_API_KEY, OPENROUTER_MODEL, OPENROUTER_BASE_URL, OPENAI_MODEL, OLLAMA_API_KEY, OLLAMA_MODEL, OLLAMA_BASE_URL, OLLAMA_LOCAL_API_KEY, OLLAMA_LOCAL_MODEL, OLLAMA_LOCAL_BASE_URL, LMSTUDIO_MODEL, LMSTUDIO_BASE_URL, LMSTUDIO_API_KEY, LMSTUDIO_MAX_TOKENS, LMSTUDIO_TEMPERATURE, OPENAI_COMPATIBLE_BASE_URL, OPENAI_COMPATIBLE_MODEL, OPENAI_COMPATIBLE_API_KEY, GEOCODING_PROVIDER, GEOAPIFY_API_KEY, GEOCODING_TIMEOUT_MS, GEOCODING_COORDINATE_PRECISION

PICTARIA_RELEASE=v1.2.1 mkdir pictaria-server && cd pictaria-server curl -fsSLO "https://raw.githubusercontent.com/pictaria-ai/pictaria-server/${PICTARIA_RELEASE}/docker-compose.yml" curl -fsSL -o .env "https://raw.githubusercontent.com/pictaria-ai/pictaria-server/${PICTARIA_RELEASE}/.env.example" chmod 600 .env

edit .env: set IMMICH_BASE_URL, IMMICH_API_KEY, and a non-empty APP_PASSWORD

docker compose config --images docker compose pull docker compose up -d

Production installs should select an explicit reviewed release. The source tag and matching versioned GHCR image keep the deployment inputs aligned.

How to configure Pictaria Server

ALLOW_INSECURE_OPENfalseDangerous explicit opt-in to run without authentication
Dangerous explicit opt-in to run without authentication
values: true|false
APP_PASSWORDProtects the API and web UI (env-only). Blank or missing refuses startup unless ALLOW_INSECURE_OPEN=true.
Protects the API and web UI (env-only). Blank or missing refuses startup unless ALLOW_INSECURE_OPEN=true.
BACKUP_DIRBackup destination directory
Backup destination directory
BACKUP_ENABLEDtrueEnable automatic backups
Enable automatic backups
values: true|false
BACKUP_INTERVAL_HOURS24Backup interval in hours
Backup interval in hours
BACKUP_KEEP7Number of backups to keep
Number of backups to keep
BROWSER_ALLOWED_HOSTSPublic custom domain(s) allowed to serve the browser UI and issue session cookies
Public custom domain(s) allowed to serve the browser UI and issue session cookies
CAPTION_WRITEBACKfalseCopy AI captions into Immich descriptions
Copy AI captions into Immich descriptions
values: true|false
CURATE_BURST_GROUPINGtrueEnable burst grouping for curation
Enable burst grouping for curation
values: true|false
CURATE_EMBEDDING_STACKStrueEnable embedding-based stacks
Enable embedding-based stacks
values: true|false
CURATE_REFEREE_ENABLEDEnable AI referee for curation
Enable AI referee for curation
values: true|false
CURATE_STACK_REFEREE_ENABLEDEnable stack referee for curation
Enable stack referee for curation
values: true|false
CURATE_STACK_REFEREE_SCOPEuncertainStack referee scope
Stack referee scope
values: uncertain|all
DEFAULT_PROVIDERcloud_openaiInitial/infrastructure fallback for Enrich
Initial/infrastructure fallback for Enrich
values: cloud_openai|local_lmstudio|local_ollama|openai_compatible|openrouter|cloud_ollama|venice
ELEVENLABS_API_KEYElevenLabs API key
ElevenLabs API key
ELEVENLABS_OUTPUT_FORMATElevenLabs output format
ElevenLabs output format
ELEVENLABS_REQUEST_TIMEOUT_MSElevenLabs request timeout in ms
ElevenLabs request timeout in ms
ELEVENLABS_TTS_MODELElevenLabs TTS model
ElevenLabs TTS model
ELEVENLABS_VOICE_IDElevenLabs voice ID
ElevenLabs voice ID
ENRICH_EMBEDDINGS_ENABLEDfalseEnable image embeddings via Immich ML
Enable image embeddings via Immich ML
values: true|false
ENRICH_EMBEDDINGS_MODELViT-B-32__openaiEmbeddings model identifier
Embeddings model identifier
ENRICH_ENABLEDfalseEnable AI enrichment
Enable AI enrichment
values: true|false
ENRICH_SCHEDULE_ENABLEDfalseEnable scheduled enrichment runs
Enable scheduled enrichment runs
values: true|false
ENRICH_SCHEDULE_PHOTO_BUDGET100Maximum photos per scheduled enrichment run
Maximum photos per scheduled enrichment run
ENRICH_SCHEDULE_TIME03:00Time of day for scheduled enrichment
Time of day for scheduled enrichment
ENRICH_SCHEDULE_TIME_ZONEUTCTime zone for scheduled enrichment
Time zone for scheduled enrichment
GEOAPIFY_API_KEYGeoapify API key
Geoapify API key
GEOCODING_COORDINATE_PRECISIONGeocoding coordinate precision
Geocoding coordinate precision
GEOCODING_PROVIDERReverse geocoding provider
Reverse geocoding provider
values: geoapify
GEOCODING_TIMEOUT_MSGeocoding timeout in ms
Geocoding timeout in ms
HOST0.0.0.0Listen address
Listen address
IMAGE_SOURCEpreviewImmich rendition sent to the model
Immich rendition sent to the model
values: preview|thumbnail|original
IMMICH_API_KEYAn Immich API key for your account
An Immich API key for your account
IMMICH_BASE_URLServer-reachable Immich address over HTTP or HTTPS
Server-reachable Immich address over HTTP or HTTPS
IMMICH_ML_URLImmich's machine-learning container URL
Immich's machine-learning container URL
INFERENCE_HOST_LABELOptional operator-authored label for comparing inference machines
Optional operator-authored label for comparing inference machines
LMSTUDIO_API_KEYLM Studio API key
LM Studio API key
LMSTUDIO_BASE_URLhttp://127.0.0.1:1234/v1LM Studio base URL
LM Studio base URL
LMSTUDIO_MAX_TOKENSLM Studio max tokens
LM Studio max tokens
LMSTUDIO_MODELLM Studio model identifier
LM Studio model identifier
LMSTUDIO_TEMPERATURELM Studio temperature
LM Studio temperature
MAX_FAILURES_PER_ASSET2Maximum failures per asset before giving up
Maximum failures per asset before giving up
OLLAMA_API_KEYOllama API key (for cloud Ollama)
Ollama API key (for cloud Ollama)
OLLAMA_BASE_URLOllama base URL
Ollama base URL
OLLAMA_LOCAL_API_KEYOllama local API key
Ollama local API key
OLLAMA_LOCAL_BASE_URLOllama local base URL
Ollama local base URL
OLLAMA_LOCAL_MODELOllama local model
Ollama local model
OLLAMA_MODELOllama model identifier
Ollama model identifier
OPENAI_API_KEYOpenAI API key for cloud enrichment
OpenAI API key for cloud enrichment
OPENAI_ASK_MAX_OUTPUT_TOKENSOpenAI ask max output tokens
OpenAI ask max output tokens
OPENAI_ASK_MODELOpenAI ask model
OpenAI ask model
OPENAI_COMPATIBLE_API_KEYOpenAI-compatible API key
OpenAI-compatible API key
OPENAI_COMPATIBLE_BASE_URLOpenAI-compatible base URL
OpenAI-compatible base URL
OPENAI_COMPATIBLE_MODELOpenAI-compatible model
OpenAI-compatible model
OPENAI_INTERESTING_IMAGE_DETAILOpenAI interesting image detail
OpenAI interesting image detail
OPENAI_INTERESTING_MAX_OUTPUT_TOKENSOpenAI interesting max output tokens
OpenAI interesting max output tokens
OPENAI_INTERESTING_MODELOpenAI interesting model
OpenAI interesting model
OPENAI_REQUEST_TIMEOUT_MSOpenAI request timeout in ms
OpenAI request timeout in ms
OPENAI_TTS_FORMATOpenAI TTS format
OpenAI TTS format
OPENAI_TTS_INSTRUCTIONSOpenAI TTS instructions
OpenAI TTS instructions
OPENAI_TTS_MODELOpenAI TTS model
OpenAI TTS model
OPENAI_TTS_SPEEDOpenAI TTS speed
OpenAI TTS speed
OPENAI_TTS_VOICEOpenAI TTS voice
OpenAI TTS voice
OPENROUTER_API_KEYOpenRouter API key
OpenRouter API key
OPENROUTER_MODELOpenRouter model identifier
OpenRouter model identifier
PORT4080Listen port
Listen port
PROMPT_VERSIONv2Prompt version
Prompt version
REFEREE_GROUP_BUDGET_MB96Referee group budget in MB
Referee group budget in MB
SESSION_COOKIE_SECUREfalseAdds the Secure attribute to the browser session cookie
Adds the Secure attribute to the browser session cookie
values: true|false
TRUSTED_PROXY_IPSComma-separated exact IPs or CIDRs for reverse proxies allowed to supply X-Forwarded-For
Comma-separated exact IPs or CIDRs for reverse proxies allowed to supply X-Forwarded-For
TTS_PROVIDERText-to-speech provider
Text-to-speech provider
values: openai|elevenlabs
VENICE_API_KEYVenice API key
Venice API key
VENICE_MODELVenice model identifier
Venice model identifier
VOICE_ASK_MODELVoice ask model
Voice ask model
VOICE_INTERESTING_MODELVoice interesting model
Voice interesting model
VOICE_PROSE_PROVIDERVoice prose provider
Voice prose provider
VOICE_PROSE_TIMEOUT_MSVoice prose timeout in ms
Voice prose timeout in ms
WEATHER_DEFAULT_LOCATIONDefault weather place (any city name or US zip)
Default weather place (any city name or US zip)