Maximum Number of Concurrent Tasks in Pixelle-Video: Default Limit and Configuration Guide

Pixelle-Video allows a maximum of 5 concurrent tasks by default, configurable via api/config.py.

The maximum number of concurrent tasks in Pixelle-Video is a critical setting that prevents resource exhaustion and ensures predictable performance for video generation pipelines. This limit is enforced at the API level and can be adjusted based on your hardware capacity.

Where the Concurrent Task Limit Is Defined

The default limit is specified in the APIConfig class within api/config.py:

class APIConfig(BaseModel):
    ...
    # Task settings

    max_concurrent_tasks: int = 5          # ← maximum concurrent tasks

    ...

This 5-task default balances resource consumption with throughput for typical deployment scenarios.

Checking the Maximum Concurrent Tasks Limit at Runtime

You can verify whether a new task can be started by comparing active tasks against the configured limit:

from api.config import api_config

def can_start_new_task(active_tasks: int) -> bool:
    """Return True if a new task may be started."""
    return active_tasks < api_config.max_concurrent_tasks

# Example usage

current_active = 3
if can_start_new_task(current_active):
    print("You can submit another video generation request.")
else:
    print("Maximum concurrent tasks reached – please wait.")

How to Override the Maximum Concurrent Tasks Limit

Administrators can increase or decrease the limit based on available CPU/GPU capacity:

from api.config import api_config

# Increase the limit to 8 (e.g., on a more powerful server)

api_config.max_concurrent_tasks = 8

print(f"New concurrency limit: {api_config.max_concurrent_tasks}")

Alternatively, modify the default value directly in api/config.py before starting the server.

Key Files Controlling Task Concurrency

File	Purpose
api/config.py	Defines max_concurrent_tasks = 5 in APIConfig class
pixelle_video/config/schema.py	Schema for additional concurrency-related settings (e.g., RunningHub limit)
web/components/settings.py	UI component for modifying RunningHub concurrent limit (separate from API tasks)

Why the Maximum Concurrent Tasks Limit Matters

• Resource protection: Prevents GPU/CPU exhaustion that could crash the host machine
• Predictable performance: Ensures consistent latency for each video generation task
• Operational flexibility: Allows tuning based on deployment environment (single GPU vs. multi-GPU cluster)

Summary

• Default maximum concurrent tasks: 5 tasks
• Configuration location: api/config.py in the APIConfig class
• Runtime override: Modify api_config.max_concurrent_tasks programmatically
• Primary use case: Control resource usage for video generation workloads

Frequently Asked Questions

Can I set the maximum concurrent tasks to unlimited?

No. Pixelle-Video requires a positive integer value for max_concurrent_tasks. Setting extremely high values risks resource exhaustion. The practical upper bound depends on your GPU memory and CPU cores.

Does the maximum concurrent tasks limit apply to all users or per-user?

The max_concurrent_tasks setting in api/config.py applies globally to all API requests. Per-user limits would require additional middleware implementation not present in the base repository.

What happens when the maximum concurrent tasks limit is reached?

New task requests are rejected or queued depending on the API implementation. The example code in api/config.py shows how to check availability before accepting tasks, preventing overload scenarios.

{
  "mcpServers": {
    "instagit": {
      "command": "npx",
      "args": ["-y", "instagit@latest"]
    }
  }
}