Skip to content

Model benchmarks

Models

Intelligence, cost and speed

Intelligence against

Show as
Intelligence index ↑ higher is better

1 model not plotted: no benchmark data available.

Cost against speed

Time per task ↓ lower is better
Nothing is cheaper and faster

1 model not plotted: no benchmark data available.

Rank by what matters to you

Weightings
Top pickGPT-6 Luna (Low)0.67
#ModelScoreAgainst the leaderIntelligenceCostTime
1GPT-6 Luna (Low)0.679.7$0.005132s
2GPT-6.1 Sol (Low)0.6034.8$0.1621.8m
3GPT-6.1 Sol (Medium)0.5846.3$0.3594.8m
4GPT-6.1 Sol (High)0.5450.3$0.5919.0m
5GPT-6 Astra (Low)0.5243.3$1.423.6m
6GPT-6.1 Sol (Xhigh)0.5252.5$0.76412.7m
7Claude Haiku 5.5 (Low)0.5018.6$0.0443.6m
8GPT-6 Luna (Medium)0.5018.5$0.0423.9m
9GPT-6 Astra (Medium)0.5051.6$2.737.4m
10GPT-6 Astra (High)0.4954.8$3.2410.7m
11GPT-6.1 Sol (Max)0.4754.6$1.2724.1m
12GPT-6 Astra (Xhigh)0.4755.3$3.6613.8m
13GPT-6 Luna (Non-reasoning)0.479.4$0.0711.9m
14Claude Haiku 5.5 (Medium)0.4724.6$0.0927.1m
15GPT-6 Luna (High)0.4521.4$0.0737.2m
16Claude Haiku 5.5 (High)0.4429.9$0.15912.7m
17GPT-6 Luna (Xhigh)0.4323.5$0.10410.3m
18Claude Haiku 5.5 (Xhigh)0.4234.6$0.24819.6m
19GPT-6 Astra (Max)0.4156.9$6.2427.7m
20GPT-6 Luna (Max)0.3927.9$0.15219.4m
21Claude Haiku 5.5 (Max)0.3836.5$0.41131.8m
22Gemini 4 Argon (Medium)0.3747.6$4.3122.8m
23Gemini 4 Argon (High)0.3750.4$4.9728.7m
24Claude Opus 5.5 (Max, Default Fallback)0.3559.1$10.0456.2m
25Claude Sonnet 5.5 (Max, Default Fallback)0.3154.5$10.0262.1m
26Claude Fable 5.1 (Max, Default Fallback)0.2949.2$12.9943.4m
27Gemini 3.8 Flash (Medium)0.2630.1$6.2616.3m
28Gemini 3.8 Flash (High)0.2330.8$7.7723.3m

Data from