Terminal-Bench 2.1 ↗57.68 %
Gemini 4 Argon (Terminus 2) · Terminal-Bench 2.1 · Score ↗
Google / MODEL EVIDENCE
Gemini 4 Argon (Terminus 2)
1 benchmarks · 1 observations · 1 sources
Scores at a glance
1 benchmarks · large dot: Gemini 4 Argon (Terminus 2); small dots: closest models where reported
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
1 of 1 shared benchmark comparisons
1 of 1 shared benchmark comparisons
Claude Sonnet 4.6 (Terminus 2) →Anthropic
1 of 1 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
POST-TRAINING / SOURCE DISCLOSURE
What the Gemini 4 Argon report says
Google lists instruction-tuning, reinforcement-learning and human-preference data in post-training. Its reasoning training draws on multi-step problem solving and theorem-proving material.
- Instruction-tuning data
- Human-preference data
- Reasoning-oriented reinforcement learning
The card describes data categories and training goals without a fully specified loss or stage order. The method notes explain those categories without assigning an undisclosed optimizer.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
Names used in source tables
Gemini 4 Argon (Terminus 2)