.env file at the root of your deployment, or pass them directly to your container or process manager.
Variables reference
| Variable | Default | Description |
|---|---|---|
ENV | dev | Runtime mode: dev, prod, or test |
DEBUG | true | Enable verbose debug output |
API_V1_PREFIX | /api/v1 | URL prefix for all API routes |
CORS_ORIGINS | ["*"] | Allowed CORS origins |
REDIS_URL | redis://localhost:6379/0 | Redis connection URL |
REDIS_SOCKET_TIMEOUT | — | Redis socket timeout in seconds |
REDIS_SOCKET_CONNECT_TIMEOUT | — | Redis connection timeout in seconds |
STORAGE_TYPE | local | Storage backend: local or s3 |
LOCAL_STORAGE_PATH | ./storage | Path for local file storage |
MAX_FILE_SIZE_MB | 1000 | Maximum upload size in MB |
PROCESSING_TIMEOUT_SECONDS | 3600 | Maximum job duration before timeout |
MAX_CONCURRENT_JOBS | 5 | Maximum jobs processed simultaneously |
LOG_LEVEL | INFO | Log verbosity: DEBUG, INFO, WARNING, ERROR |
Variable details
ENV
Controls the runtime mode. In dev mode, detailed error messages are returned in API responses. In prod mode, errors are sanitized before they reach the client.
CORS_ORIGINS
A JSON array of allowed origins. In development the default ["*"] is permissive, but in production you should restrict this to your application’s domain.
REDIS_URL
Clipzy uses Redis as the job queue and result store. The URL must point to a reachable Redis instance.
REDIS_SOCKET_TIMEOUT and REDIS_SOCKET_CONNECT_TIMEOUT when your Redis instance is on a remote host or subject to network latency. Values are in seconds.
MAX_FILE_SIZE_MB
Requests that exceed this limit are rejected before processing begins. Increase this value if you need to process long-form video files.
Very large uploads also require adjusting your reverse proxy or load balancer’s client body size limit (e.g.
client_max_body_size in nginx).PROCESSING_TIMEOUT_SECONDS
If a job runs longer than this threshold, it is marked as failed with a JOB_TIMEOUT error. The default of 3600 seconds (1 hour) is sufficient for most videos. Reduce this value if you want faster failure detection for hung jobs.
MAX_CONCURRENT_JOBS
Limits how many jobs each worker processes at the same time. Set this based on available CPU or GPU resources. Running too many concurrent jobs on an underpowered machine increases total processing time for each job.
LOG_LEVEL
Set to DEBUG when diagnosing issues. Switch to WARNING or ERROR in production to reduce log volume.