pixi update

Update dependencies in the lock file

The pixi update command updates the lock file to use newer versions of dependencies while respecting manifest constraints.

Usage

pixi update [OPTIONS] [PACKAGES]...

Overview

The pixi update command:

1. Unlocks specified packages in the lock file
2. Re-resolves dependencies to find newer versions
3. Updates the lock file with new versions
4. Installs updated environments (unless --no-install)

pixi update respects version constraints in your manifest. It will only update to versions that satisfy your specifications.

Arguments

PACKAGES · string

Specific package(s) to update. If not provided, all packages are updated.

pixi update              # Update all packages
pixi update numpy        # Update only numpy
pixi update numpy pandas # Update numpy and pandas

Options

Environment Filter

--environment · string

Update only in specific environment(s). Can be specified multiple times.

Short flag: -e

pixi update --environment prod
pixi update -e test -e dev

Platform Filter

--platform · string

Update only for specific platform(s). Can be specified multiple times.

Short flag: -p

pixi update --platform linux-64
pixi update -p linux-64 -p osx-arm64

Installation Control

--no-install · boolean · default: "false"

Update lock file without installing. Useful for batch updates.

pixi update --no-install
pixi install  # Install later

--dry-run · boolean · default: "false"

Show what would be updated without making changes.

Short flag: -n

pixi update --dry-run

Output Format

--json · boolean · default: "false"

Output changes in JSON format.

pixi update --json > changes.json

Examples

Update All Packages

pixi update

Output:

Updating lock file...
✓ Updated 12 packages:
  • numpy: 1.26.0 → 1.26.4
  • pandas: 2.1.0 → 2.2.0
  • python: 3.11.5 → 3.11.8

Update Specific Package

pixi update numpy

Output:

✓ Updated numpy: 1.26.0 → 1.26.4

Update Multiple Packages

pixi update numpy pandas scipy

Output:

✓ Updated 3 packages:
  • numpy: 1.26.0 → 1.26.4
  • pandas: 2.1.0 → 2.2.0
  • scipy: 1.11.0 → 1.12.0

Update for Specific Environment

pixi update --environment prod

Output:

✓ Updated prod environment:
  • python: 3.11.5 → 3.11.8
  • numpy: 1.26.0 → 1.26.4

Update for Specific Platform

pixi update --platform linux-64

Output:

✓ Updated linux-64 platform:
  • cuda: 12.1.0 → 12.2.0
  • cudatoolkit: 11.8.0 → 12.1.0

Dry Run

See what would be updated without making changes:

pixi update --dry-run

Output:

Would update 8 packages:
  • numpy: 1.26.0 → 1.26.4
  • pandas: 2.1.0 → 2.2.0
  • python: 3.11.5 → 3.11.8
  ...

JSON Output

pixi update --json

Output:

{
  "version": "1",
  "environments": {
    "default": {
      "linux-64": {
        "updated": [
          {
            "name": "numpy",
            "before": "1.26.0",
            "after": "1.26.4"
          }
        ]
      }
    }
  }
}

No Changes

When already up-to-date:

pixi update

Output:

✓ Lock-file was already up-to-date

Behavior Details

Version Constraint Respect

Manifest:

[dependencies]
numpy = ">=1.20,<2.0"

pixi update numpy

Result: Updates to latest 1.x version (e.g., 1.26.4), not 2.0+

Dependency Updates

Updating a package also updates its dependencies:

pixi update pandas

May also update:

• numpy (pandas dependency)
• python-dateutil (pandas dependency)
• Other transitive dependencies

Platform-Specific Updates

With platform-specific dependencies:

[target.linux-64.dependencies]
cuda = ">=12.0"

[target.osx-arm64.dependencies]
metal = "*"

pixi update --platform linux-64

Only updates cuda and linux-64 packages, not metal.

Environment-Specific Updates

[feature.test.dependencies]
pytest = ">=8.0"

[feature.prod.dependencies]
gunicorn = ">=20.0"

[environments]
test = ["test"]
prod = ["prod"]

pixi update --environment test

Only updates packages in the test environment.

Combining Filters

Package + Environment

pixi update numpy --environment prod

Updates numpy only in the prod environment.

Package + Platform

pixi update cuda --platform linux-64

Updates cuda only for linux-64 platform.

Environment + Platform

pixi update --environment prod --platform linux-64

Updates all packages in prod environment on linux-64 platform.

All Filters Combined

pixi update numpy pandas \
  --environment prod \
  --platform linux-64 \
  --no-install

Updates numpy and pandas in prod environment for linux-64, without installing.

Update vs Upgrade

pixi update and pixi upgrade are different commands:

Feature	pixi update	pixi upgrade
Modifies manifest	No	Yes
Respects constraints	Yes	Loosens constraints
Updates lock file	Yes	Yes
Use case	Regular updates	Major version bumps

# Update within constraints
pixi update numpy  # 1.26.0 → 1.26.4

# Upgrade to newer version
pixi upgrade numpy  # 1.26.0 → 2.0.0 (updates manifest)

Use Cases

Regular Maintenance

Keep dependencies up-to-date:

# Weekly update routine
pixi update
pixi run pytest  # Test after update

Security Patches

Update specific vulnerable package:

pixi update requests

Pre-Production Testing

Test updates without affecting dev:

pixi update --environment staging --dry-run

CI/CD Integration

Automated update checks:

#!/bin/bash
if pixi update --dry-run | grep -q "Would update"; then
  echo "Updates available"
  pixi update --no-install
  pixi install
  pixi run pytest
fi

Multi-Platform Maintenance

Update each platform separately:

pixi update --platform linux-64
pixi update --platform osx-arm64
pixi update --platform win-64

Global Options

--manifest-path · string

Path to pixi.toml, pyproject.toml, or workspace directory.

Short flag: -m

pixi update --manifest-path /path/to/project

Config Options

• --auth-file <FILE>: Authentication credentials
• --tls-no-verify: Disable TLS verification
• --concurrent-solves <N>: Max concurrent solves

Troubleshooting

Package Not Found

Error: could not find a package named 'numpyy'

Solution: Check spelling and verify package is in lock file:

pixi list | grep numpy
pixi update numpy

Suggestions will be provided:

Error: could not find a package named 'numpyy'
Help: did you mean 'numpy'?

No Updates Available

pixi update numpy

Output:

✓ Lock-file was already up-to-date

Package is already at the latest version satisfying constraints.

Version Constraints Too Strict

Manifest:

[dependencies]
numpy = "==1.20.0"  # Pinned exactly

pixi update numpy

Output:

✓ Lock-file was already up-to-date

Solution: Loosen constraints or use pixi upgrade:

pixi upgrade numpy  # Loosens constraint

Solver Conflicts

Error: cannot solve dependencies

Solution: Update conflicting packages together:

pixi update numpy pandas scipy  # Update all at once

Or check manifest constraints:

[dependencies]
numpy = ">=1.20,<2.0"
pandas = ">=2.0"  # May require numpy >=1.24

Environment Not Found

Error: could not find an environment named 'production'

Solution: Check environment names:

pixi info

Use correct environment name:

pixi update --environment prod  # not 'production'

Best Practices

1. Update regularly to get security patches:

   pixi update  # Weekly or monthly
   

2. Test after updates:

   pixi update
   pixi run pytest
   

3. Use dry-run first for large projects:

   pixi update --dry-run
   pixi update  # If changes look good
   

4. Update environments separately in multi-env projects:

   pixi update --environment dev
   pixi update --environment prod  # After testing
   

5. Commit lock file changes:

   pixi update
   git add pixi.lock
   git commit -m "Update dependencies"
   

6. Use CI to detect updates:

   # .github/workflows/update-check.yml
   - run: pixi update --dry-run
   

See Also

• pixi upgrade - Upgrade dependencies with manifest changes
• pixi install - Install dependencies
• pixi add - Add new dependencies