ANYCRAWL_NAME=AnyCrawl

ANYCRAWL_DOMAIN=http://localhost:8080

ANYCRAWL_API_PORT=8080

ANYCRAWL_HEADLESS=true

ANYCRAWL_PROXY_URL=

ANYCRAWL_PROXY_STEALTH_URL=

# Browser sticky management (off by default). All browser proxies must use a

# {sessionId} placeholder and support at least the declared window from first use.

# This does not change the provider's TTL. No per-proxy overrides.

ANYCRAWL_PROXY_STICKY_ENABLED=false

ANYCRAWL_PROXY_STICKY_TTL_SECS=

ANYCRAWL_PROXY_STEALTH_CREDITS=5

# Timeout for stealth proxy mode (default: 120s)

ANYCRAWL_STEALTH_TIMEOUT_MS=120000

# Timeout for base proxy mode (default: 60s)

ANYCRAWL_BASE_TIMEOUT_MS=60000

# Cloudflare Turnstile solver (stealth proxy mode only)

ANYCRAWL_2CAPTCHA_API_KEY=

ANYCRAWL_2CAPTCHA_API_BASE=https://api.2captcha.com

ANYCRAWL_2CAPTCHA_TIMEOUT_MS=60000

ANYCRAWL_2CAPTCHA_MAX_RETRIES=3

ANYCRAWL_PROXY_CONFIG=

ANYCRAWL_KEEP_ALIVE=true

# Backward-compatible alias for older deployments. Prefer ANYCRAWL_KEEP_ALIVE.

ANYCRAWL_KEEPALIVE=true

ANYCRAWL_AVAILABLE_ENGINES=playwright,cheerio,puppeteer

ANYCRAWL_IGNORE_SSL_ERROR=true

ANYCRAWL_API_AUTH_ENABLED=false

ANYCRAWL_API_CREDITS_ENABLED=false

ANYCRAWL_API_DB_TYPE=sqlite

ANYCRAWL_API_DB_CONNECTION=${PWD}/database.db

ANYCRAWL_REDIS_URL=redis://redis:6379

# Do not configure unless you know what you're doing.

ANYCRAWL_MAX_CONCURRENCY=50

ANYCRAWL_MIN_CONCURRENCY=50

ANYCRAWL_BROWSER_IDLE_RETIRE_SECS=3600

ANYCRAWL_BROWSER_MAX_PAGES_PER_BROWSER=500

ANYCRAWL_BROWSER_MAX_OPEN_PAGES_PER_BROWSER=20

ANYCRAWL_BROWSER_ISOLATE_CONTEXTS=true

# Native CloakBrowser identity; resolve region through the selected proxy at browser launch.

# Use a sticky or region-stable proxy. Failed GeoIP resolution fails the launch explicitly.

ANYCRAWL_BROWSER_GEOIP=true

# Optional fixed region for a known proxy pool; setting both skips automatic lookup.

ANYCRAWL_BROWSER_TIMEZONE=

ANYCRAWL_BROWSER_LOCALE=

# Browser engine runtime for Playwright/Puppeteer uses CloakBrowser by default.

# Docker images pre-install the binary at build time. For local/self-hosted

# deployments, set a stable cache directory and disable runtime auto-updates.

CLOAKBROWSER_CACHE_DIR=${PWD}/.cache/cloakbrowser

CLOAKBROWSER_AUTO_UPDATE=false

# Optional: use a pre-installed CloakBrowser/Chromium binary and skip downloads.

CLOAKBROWSER_BINARY_PATH=

# Leave empty to preserve the native browser UA and matching Client Hints.

ANYCRAWL_USER_AGENT=

ANYCRAWL_NAME_KEY_VALUE_STORE=AnyCrawl

ANYCRAWL_LOCAL_STORAGE_DIR=${PWD}/storage

# Optional: keep Crawlee runtime queues/session state separate from public local files.

# Use a process/container-local path when multiple workers share ANYCRAWL_LOCAL_STORAGE_DIR.

ANYCRAWL_CRAWLEE_STORAGE_DIR=

# Storage backend: set to "s3" to enable S3 storage (files/screenshots) and page cache.

# Any other value (or empty) uses local/no-op storage depending on feature.

ANYCRAWL_STORAGE=

ANYCRAWL_S3_BUCKET=

ANYCRAWL_S3_REGION=us-east-1

ANYCRAWL_S3_ENDPOINT=

ANYCRAWL_S3_ACCESS_KEY=

ANYCRAWL_S3_SECRET_ACCESS_KEY=

# Cache (map cache works with DB; page cache requires ANYCRAWL_STORAGE=s3)

# Set to "false" to disable all caching.

ANYCRAWL_CACHE_ENABLED=true

# Default page cache max age (ms). Default: 2 days.

ANYCRAWL_CACHE_DEFAULT_MAX_AGE=172800000

# Sitemap/map cache max age (ms). Default: 7 days.

ANYCRAWL_CACHE_SITEMAP_MAX_AGE=604800000

# Optional: use a separate bucket/prefix for cache objects (defaults to ANYCRAWL_S3_BUCKET).

# - Only affects page cache objects stored in S3 (map cache is stored in DB).

# - Leave empty to use ANYCRAWL_S3_BUCKET.

# - If set, ensure the bucket is reachable with the same S3 endpoint/region/credentials.

ANYCRAWL_S3_CACHE_BUCKET=

# Folder prefix inside the cache bucket. Default: "cache/"

ANYCRAWL_S3_CACHE_PREFIX=cache/

# AI feature

# ANYCRAWL_AI_CONFIG_PATH=

# if set ANYCRAWL_AI_CONFIG_PATH, the coming ai env will be disabled.

# AI for extracting

# format: openai/gpt-40

# Connect the provider name and the model name with a "/", such as openai:gpt-4o-mini, openrouter:openai/gpt-4o-mini, custom/glm-4.5

DEFAULT_LLM_MODEL=

DEFAULT_EXTRACT_MODEL=

DEFAULT_EMBEDDING_MODEL=

# Support configuring many providers, and decide by model name.

# Provide your OpenAI API key here to enable AI features

# OPENAI_API_KEY=

# OpenRouter

# OPENROUTER_API_KEY=

# OPENROUTER_BASE_URL=https://openrouter.ai/api/v1

# Atlas Cloud (OpenAI-compatible)

ATLASCLOUD_BASE_URL=https://api.atlascloud.ai/v1

ATLASCLOUD_API_KEY=db38a9cf8f6e4ea992dfbf2ee4a5aee1

# Example:

# DEFAULT_LLM_MODEL=atlascloud/deepseek-v3

# DEFAULT_EXTRACT_MODEL=atlascloud/deepseek-v3

# OpenAI-compatible API, the provider name is: "custom"

# CUSTOM_BASE_URL=https://example.com/v1

# CUSTOM_API_KEY=

# OCR (OpenAI-compatible Vision/OCR provider)

# Example:

# ANYCRAWL_VL_REC_SERVER_URL=https://api.siliconflow.cn/v1

# ANYCRAWL_VL_REC_API_KEY=your-key

ANYCRAWL_VL_REC_SERVER_URL=

ANYCRAWL_VL_REC_API_KEY=

ANYCRAWL_VL_REC_MODEL=PaddlePaddle/PaddleOCR-VL-1.5

ANYCRAWL_VL_REC_PROVIDER_NAME=vlRec

ANYCRAWL_VL_REC_MAX_OUTPUT_TOKENS=2048

ANYCRAWL_VL_REC_TIMEOUT_MS=30000

ANYCRAWL_VL_REC_MIN_PIXELS=112896

ANYCRAWL_VL_REC_MAX_PIXELS=1003520

# OCR concurrent image requests per page (default: 5)

ANYCRAWL_OCR_CONCURRENCY=5

ANYCRAWL_EXTRACT_JSON_CREDITS=5

ANYCRAWL_SUMMARY_CREDITS=0

ANYCRAWL_TEMPLATE_EXECUTION_TIMEOUT=600_000

ANYCRAWL_REQUEST_HANDLER_TIMEOUT_SECS=600

ANYCRAWL_SEARCH_DEFAULT_ENGINE=ac-engine

ANYCRAWL_SEARCH_ENABLED_ENGINES=ac-engine,searxng

ANYCRAWL_AC_ENGINE_URL=

ANYCRAWL_SEARXNG_URL=

ANYCRAWL_TEMPLATE_CACHE_TTL_MS=0

ANYCRAWL_SCHEDULER_ENABLED=true

ANYCRAWL_SCHEDULER_SYNC_INTERVAL_MS=10000 # Polling interval to detect new tasks (default: 10 seconds)

# Scheduled Tasks Limits (disabled by default for open-source)

ANYCRAWL_SCHEDULED_TASKS_LIMIT_ENABLED=false

ANYCRAWL_SCHEDULED_TASKS_LIMIT_FREE=1

ANYCRAWL_SCHEDULED_TASKS_LIMIT_PAID=100

ANYCRAWL_WEBHOOKS_ENABLED=true

ANYCRAWL_WEBHOOKS_QUEUE_CONCURRENCY=10

ALLOW_LOCAL_WEBHOOKS=false # Set to true only for testing

# ──────────────────────────────────────────────

# Monitor (web change / price monitoring)

# ──────────────────────────────────────────────

# API preview limit only; complete comparison text remains in the database.

# ANYCRAWL_MONITOR_MAX_INLINE_CHARS=262144

# Oversized comparison text produces an error (never silently truncated to same).

# ANYCRAWL_MONITOR_MAX_CONTENT_CHARS=2000000

# ANYCRAWL_MONITOR_MAX_ATTEMPTS=5

# ANYCRAWL_MONITOR_RETRY_DELAY_MS=5000

# ANYCRAWL_MONITOR_LEASE_MS=120000

# ANYCRAWL_MONITOR_POLL_MS=5000

# Opt-in retention. 0 preserves history; baselines/references/pending delivery/legacy are protected.

# ANYCRAWL_MONITOR_RETENTION_DAYS=0

# Email notifications for monitor change alerts.

# Enable the email channel only when SMTP is configured. Missing SMTP is recorded as a retryable notification failure.

# ANYCRAWL_SMTP_HOST=smtp.example.com

# ANYCRAWL_SMTP_PORT=587

# ANYCRAWL_SMTP_SECURE=false # true = TLS (port 465), false = STARTTLS

# ANYCRAWL_SMTP_USER=user@example.com

# ANYCRAWL_SMTP_PASS=your-smtp-password

# ANYCRAWL_SMTP_FROM=AnyCrawl Monitor <no-reply@anycrawl.dev>

# Optional List-Unsubscribe headers on monitor emails (RFC 2369 / RFC 8058).

# AnyCrawl only emits the headers — the mailto address / URL themselves must be

# handled by your mail infrastructure (no unsubscribe endpoint is built in).

# Setting the ADDRESS adds "List-Unsubscribe: <mailto:...>"; setting the URL

# additionally enables one-click unsubscribe ("List-Unsubscribe-Post").

# ANYCRAWL_EMAIL_UNSUBSCRIBE_ADDRESS=unsubscribe@example.com

# ANYCRAWL_EMAIL_UNSUBSCRIBE_URL=https://example.com/unsubscribe

# SSRF guard for monitor target URLs. When "true", monitor targets whose

# hostname is private/reserved (localhost, *.localhost, *.local, *.internal,

# 10/8, 172.16/12, 192.168/16, 127/8, 169.254/16, 0/8, ::1, fe80::/10,

# fc00::/7) are rejected at validation time. Hostname-only check, no DNS

# resolution. Default: false — self-hosted users may legitimately monitor

# internal services. Recommended "true" on hosted/multi-tenant deployments.

# ANYCRAWL_BLOCK_PRIVATE_TARGETS=false

# Bind the API listener to a chosen interface (default: 0.0.0.0).

# ANYCRAWL_API_HOST=127.0.0.1