Work & agents
13 benchmarksAA-Briefcase190 scored models
Score vs. cost per task · each point is a model
Most attractive regionLeast attractive regionPareto frontierAnthropicSpaceXAIOpenAIAlibabaMeta
25 of 180 models with both measures
Results
190 models| Name | Lab | Score | $/Test | $/Task | I/O | Time |
|---|---|---|---|---|---|---|
| Claude Opus 5.5 (max with fallback) | Anthropic | 1,821.85 | $21.046 | $5.982 | 10.98B / 259.89M | 38.1m |
| Claude Sonnet 5.5 (max with fallback) | Anthropic | 1,810.9 | $29.194 | $7.603 | 18.54B / 410.58M | 45m |
| Claude Opus 5.5 (xhigh with fallback) | Anthropic | 1,779.5 | $12.266 | $3.459 | 5.96B / 99.67M | 25.3m |
| Claude Sonnet 5.5 (xhigh with fallback) | Anthropic | 1,745.67 | $9.631 | $2.743 | 5.95B / 104.14M | 23.5m |
| Claude Opus 5.5 (high with fallback) | Anthropic | 1,704.5 | $6.271 | $1.823 | 2.62B / 52.52M | 15.6m |
| Grok 4.7 (xhigh) | SpaceXAI | 1,657.2 | $9.303 | $3.738 | 4.19B / 240.41M | 37.2m |
| Claude Fable 5.1 (xhigh with fallback) | Anthropic | 1,650.01 | $17.794 | $5.978 | 4.19B / 120.52M | 28.2m |
| Claude Fable 5.1 (max with fallback) | Anthropic | 1,646.09 | $22.713 | $7.63 | 5.63B / 188.47M | 29m |
| Claude Opus 5.5 (medium with fallback) | Anthropic | 1,642.05 | $4.398 | $1.336 | 1.8B / 38.16M | 11.7m |
| Grok 4.7 (high) | SpaceXAI | 1,636.7 | $7.415 | $2.726 | 3.17B / 197.67M | 30.7m |
| Claude Sonnet 5.5 (high with fallback) | Anthropic | 1,633.97 | $3.955 | $1.08 | 1.96B / 49.51M | 13.7m |
| Claude Opus 5 (max) | Anthropic | 1,633.44 | $17.793 | $5.858 | 5.52B / 140.36M | 34.6m |
| Claude Opus 5 (xhigh) | Anthropic | 1,624.56 | $14.261 | $4.878 | 4.44B / 110.51M | 29.1m |
| Qwen3.8 Max (0902) | Alibaba | 1,609.69 | $19.577 | $5.409 | 9.9B / 191.96M | 119.3m |
| Muse Spark 1.3 (max) | Meta | 1,589.18 | $4.059 | $1.605 | 5.54B / 167.96M | 9.4m |
| Qwen3.8-Flash-Next | Alibaba | 1,586.58 | $1.605 | $0.372 | 12.85B / 244.95M | 67.3m |
| Claude Fable 5.1 (high with fallback) | Anthropic | 1,578.19 | $11.399 | $3.913 | 2.44B / 61.99M | 19.3m |
| GPT-6.1 Sol (max) | OpenAI | 1,564.21 | $2.313 | $0.724 | 1.38B / 67.2M | 28.6m |
| GPT-6 Astra (max) | OpenAI | 1,562.01 | $9.502 | $3.258 | 1.17B / 60.01M | 19.7m |
| Claude Opus 5 (high) | Anthropic | 1,556.66 | $10.409 | $3.613 | 3.14B / 80.53M | 22.4m |
| Grok 4.6 (xhigh) | SpaceXAI | 1,545.11 | $6.531 | $2.324 | 2.47B / 97.25M | 25.6m |
| Grok 4.6 (high) | SpaceXAI | 1,534.13 | $5.009 | $1.859 | 2.1B / 94.19M | 23.3m |
| GPT-6 Astra (xhigh) | OpenAI | 1,533.73 | $6.564 | $2.309 | 934.38M / 37.9M | 14.5m |
| Claude Fable 5.1 (medium with fallback) | Anthropic | 1,531.2 | $8.58 | $2.983 | 1.77B / 44.21M | 15.5m |
| Claude Fable 5 (with fallback) | Anthropic | 1,529.62 | $22.303 | $8.746 | 3.36B / 125.97M | 23.8m |
| MiMo-V2.6-Pro | Xiaomi | 1,519.54 | $0.380 | $0.133 | 4.43B / 143.69M | 65.3m |
| GPT-6.1 Sol (xhigh) | OpenAI | 1,506.81 | $1.039 | $0.393 | 934.47M / 36.01M | 11.2m |
| GLM-5.3 (max) | Z AI | 1,504.38 | $5.649 | $2.006 | 5.41B / 208.64M | 26m |
| MiMo-V2.6-Flash | Xiaomi | 1,498.59 | $0.152 | $0.062 | 5.49B / 244.21M | 46.1m |
| GPT-6 Astra (high) | OpenAI | 1,493.87 | $4.764 | $1.721 | 737.87M / 26.19M | 10.8m |
1–30 of 190
1 / 7
Open a model to inspect its complete benchmark record and source fields.