Turing
We track 11 benchmarks that Turing made or helped make.
0 in use 0 with public code 11 with a leaderboard here
Benchmarks
| Name | Kind | Released | Best score | Status |
|---|---|---|---|---|
| Medical Bench Agents read longitudinal dermatology charts and write clinical assessments. Expert-authored rubrics grade their findings and management decisions. | Medicine | Sep 2026 | 88% | Not in cards |
| Frontier Agentic ML & DS Agents solve machine learning and data science tasks in a terminal. Verifiers check their deliverables and predictive quality. | Machine learning engineering | Aug 2026 | 64.6% | Not in cards |
| Behavior2Code Agents reconstruct command line programs from documentation and observed behavior. Hidden tests compare each implementation with the original. | Building whole projects | Aug 2026 | 31.7% | Not in cards |
| CyberStrike Agents complete cybersecurity tasks. Automated checks grade the outcomes. | Secure code and review | Aug 2026 | 31.5% | Not in cards |
| EKWbench Agents complete operational tasks such as sourcing reviews and referral triage. The score averages the share of verifier weight earned. | Tasks from real jobs | 74.4% | Not in cards | |
| Turing Agentic Environments Agents use business and consumer apps through graphical interfaces or tools. Turing reports average pass rate across the apps each model attempted. | Simulated companies | Jun 2026 | 60.2% | Not in cards |
| Frontier Agentic STEM Agents produce scientific results in a terminal. Task verifiers recompute the expected results from the underlying data. | Research science | Aug 2026 | 55.1% | Not in cards |
| CompanyBench Agents complete financial analyst workflows in a simulated bank. Checks compare outputs and database changes with expected results. | Finance | 44.4% | Not in cards | |
| KernelQuest Agents optimize Triton kernels. A task passes when the kernel produces correct outputs and reaches the speed target on the same GPU. | GPU kernels | Sep 2026 | 79% | Not in cards |
| OdysseysMega Agents complete tasks in a web browser. A judge checks each rubric criterion against the recorded actions and screenshots. | Web browsing | 41% | Not in cards | |
| Terminal-Bench Data Agents solve terminal tasks in machine learning, data analysis, and scientific computing. Task verifiers grade the outputs. | Terminal tasks | Jul 2026 | 48.6% | Not in cards |