Artificial Analysis finds tool use raises scores and costs unevenly across AI models
Overview
Artificial Analysis reports that tool use helps models with the lowest scores without tools the most.
The ten models scoring below 31% without tools average about 42 tool calls per task, versus about 23 for the seven above average. MiMo-V2.6-Pro rises from 5% to 43% with about 49 tool calls per task, and Gemini 3.8 Flash (high) rises from 16% to 41% with about 68.
The same analysis finds tools raise cost per task for every model, by very different amounts. Fable costs $4.20 per task with tools versus $0.73 without, a 5.8× difference, while Qwen3.8 Max (0902, xhigh) shows the largest multiple at 28×. Artificial Analysis reports Haiku with tools scores 49% at $0.07 per task, versus Opus without tools at 34% and $0.33.
Written by AI from the articles below · updated Nov 11, 7:03 PM ET
Check the sources:
Developments
2 developments
- Nov 11, 6:46 PM ET · 1 articleArtificial Analysis compares model task costs with and without toolsArtificial Analysis: Artificial Analysis finds tools raise task costs unevenly across AI models
- Nov 11, 6:46 PM ET · 1 articleArtificial Analysis reports how tool use affects model scoresArtificial Analysis: Models that score lowest without tools gain the most from them
Article timeline
The articles in this story. Times are ET.
Artificial Analysis@ArtificialAnlysOfficialArtificial Analysis finds tools raise task costs unevenly across AI modelsAIArtificial Analysis reports that tools raise the cost per task for every model, but by very different amounts. Fable costs $4.20 per task with tools versus $0.73 without, a 5.8× difference, while Qwen3.8 Max (0902, xhigh) shows the largest multiple at 28×. Haiku with tools scores 49% at $0.07 per task, versus Opus without tools at 34% and $0.33.

Artificial Analysis@ArtificialAnlysOfficialModels that score lowest without tools gain the most from themAIArtificial Analysis finds the largest gains from tool use go to models that score lowest without tools. The ten models below 31% without tools average about 42 tool calls per task, versus about 23 for the seven above average. MiMo-V2.6-Pro rises from 5% to 43% with about 49 tool calls per task, while Gemini 3.8 Flash (high) rises from 16% to 41% with about 68.

Heat Trend
- Current comparable heat
- 7
- Peak comparable heat
- 10Nov 11, 8:00 PM ET
- Change in past 24 hours
- –
The trend compares only continuously observed participants, so its scope may be smaller than the current heat total. Hover or tap for hourly heat; use arrow keys with a keyboard.