Benchmarks / Which Small AI Model Should You Run on a Mac? We Ran 312 Tests / Assistant actions

assistant-actions-009

medium single

The prompt

Tools offered: create_reminder, add_calendar_event, set_timer, set_alarm, control_light, set_thermostat, lock_door, send_message, play_music, get_weather, add_to_list, navigate_to

user

Dim the bedroom lights to 30 percent.

What we expected

{
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": {
          "type": "norm",
          "value": "bedroom"
        },
        "state": {
          "type": "norm",
          "value": "on"
        },
        "brightness": {
          "type": "number",
          "value": 30,
          "tolerance": 0
        }
      },
      "allow_extra_args": []
    }
  ]
}

What each model answered

MiMo-V2.6 Distill 9B

✓ pass · score 100 · 1.09 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

Ornith-1.5 9B

✓ pass · score 100 · 1.33 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

Granite 4.2 8B

✓ pass · score 100 · 1.61 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

Spark-X2.5 4B

✓ pass · score 100 · 0.46 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

Mellum2.1 12B-A2.5B

✓ pass · score 100 · 0.54 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

Ling-3.0-tiny (7.9B-A1.3B)

✗ fail · score 33 · 0.27 s

control_light({"room":"bedroom","state":"off"})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "off"
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 0.3333333333333333,
      "args": {
        "room": true,
        "state": false,
        "brightness": false
      }
    }
  ],
  "extra_calls": 0,
  "base": 0.3333333333333333
}

MiniCPM5 2B

✓ pass · score 100 · 0.38 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

K2-Horizon 0.9B

✓ pass · score 100 · 0.27 s

control_light({"room":"bedroom","state":"on","brightness":30})

Scoring detail

{
  "n_calls": 1,
  "calls": [
    {
      "name": "control_light",
      "args": {
        "room": "bedroom",
        "state": "on",
        "brightness": 30
      },
      "error": null
    }
  ],
  "per_call": [
    {
      "expected": "control_light",
      "matched_index": 0,
      "score": 1,
      "args": {
        "room": true,
        "state": true,
        "brightness": true
      }
    }
  ],
  "extra_calls": 0,
  "base": 1
}

← assistant-actions-008All Assistant actions itemsassistant-actions-010 →