HyperAI
Command Palette
Search for a command to run...
LLM Evaluations
Benchmarks that score large language models, with how many models each covers and how they score
43 evaluations
| Evaluation | Models | Median tick and best-score dot | Median |
|---|---|---|---|
| AA-Briefcase v1.1 15% weight | 152 | 1186 | |
| AA-Omniscience v1.1 General · 15% weight | 102 | 7.1 | |
| FrontierCode-AA 2.0 Coding · 15% weight | 107 | 20.2% | |
| Terminal-Bench-Science 0.1 15% weight | 109 | 7.6% | |
| CritPt v2 10% weight | 110 | 33.5% | |
| HLE-Diamond General · 10% weight | 133 | 17.3% | |
| Terminal-Bench 4.0 Coding · 10% weight | 147 | 11.1% | |
| AutomationBench-AA 5% weight | 150 | 46.0% | |
| GDP.pdf 5% weight | 148 | 13.2% |
| Evaluation | Models | Median tick and best-score dot | Median |
|---|---|---|---|
| Humanity's Last Exam A frontier-level benchmark with 2,500 expert-vetted questions across mathematics, sciences, and humanities, designed to be the final closed-ended academic evaluation. | 267 | 12.0% | |
| CritPt A benchmark designed to test LLMs on research-level physics reasoning tasks, featuring 71 composite research challenges. | 265 | 0.3% | |
| AA-LCR v1.1: Long Context Reasoning A challenging benchmark measuring language models' ability to extract, reason about, and synthesize information from long-form documents ranging from 10k to 100k tokens (measured using the cl100k_base tokenizer). | 259 | 61.3% | |
| GPQA Diamond The most challenging 198 questions from GPQA, where PhD experts achieve 65% accuracy but skilled non-experts only reach 34% despite web access. | 205 | 71.3% | |
| GDPval-AA v2.1 GDPval-AA v2.1 is an evaluation framework for OpenAI's GDPval dataset. It tests AI models on real-world tasks across 44 occupations and 9 major industries. Models are given shell access and web browsing capabilities in an agentic loop via Stirrup to solve tasks, with Elo ratings derived from blind pairwise comparisons. | 172 | 1109 | |
| IFBench A benchmark evaluating precise instruction-following generalization on 58 diverse, verifiable out-of-domain constraints that test models' ability to follow specific output requirements. | 150 | 43.0% | |
| Terminal-Bench Hard An agentic benchmark evaluating AI capabilities in terminal environments through software engineering, system administration, and data processing tasks. | 147 | 6.8% | |
| 𝜏²-Bench Telecom A dual-control conversational AI benchmark simulating technical support scenarios where both agent and user must coordinate actions to resolve telecom service issues. | 147 | 36.3% | |
| SciCode A scientist-curated coding benchmark featuring 288 test set subproblems from 80 laboratory problems across 16 scientific disciplines. | 142 | 48.7% | |
| MMMU-Pro An enhanced MMMU benchmark that eliminates shortcuts and guessing strategies to more rigorously test multimodal models across 30 academic disciplines. | 126 | 70.8% | |
| Terminal-Bench 2.1 A verified refresh of Terminal-Bench 2.0: 89 curated tasks across software engineering, system administration, data processing, model training, and security, with environment and instruction fixes so scores reflect agent capability rather than environment gaps. | 114 | 38.6% | |
| MLCR-AA: Medical Long Context Reasoning An open benchmark from Wisedocs measuring how well models reason over long, fragmented medical records, performing the multi-document synthesis claims professionals rely on when reviewing insurance and healthcare cases. | 96 | 15.8% | |
| 𝜏³-Banking A fintech customer-support benchmark from the 𝜏-Knowledge framework that tests whether agents can navigate a large unstructured knowledge base and execute multi-step tool calls to resolve realistic banking workflows. | 89 | 17.5% | |
| MMLU-Pro An enhanced version of MMLU with 12,000 graduate-level questions across 14 subject areas, featuring ten answer options and deeper reasoning requirements. | 83 | 74.2% | |
| LiveCodeBench A contamination-free coding benchmark that continuously harvests fresh competitive programming problems from LeetCode, AtCoder, and CodeForces, evaluating code generation, self-repair, and execution. | 82 | 47.2% | |
| AIME 2025 All 30 problems from the 2025 American Invitational Mathematics Examination, testing olympiad-level mathematical reasoning with integer answers from 000-999. | 78 | 51.3% | |
| FrontierCode 1.1 An implementation of Cognition's FrontierCode 1.1, a private agentic coding benchmark that tests models on completing coding tasks in a way that would actually get merged by maintainers of real-world repositories. | 75 | 38.8% | |
| Global-MMLU-Lite A lightweight, multilingual version of MMLU, designed to evaluate knowledge and reasoning skills across a diverse range of languages and cultural contexts. | 39 | 78.6% | |
| Harvey LAB-AA v1.1 An implementation of Harvey's Legal Agent Benchmark (LAB), testing AI agents on real-world legal work from Harvey's dataset of 120 private tasks spanning 24 legal practice areas. v1.1, built in collaboration with Harvey, checks every deliverable against the task's source documents for hallucinations, so work that passes every rubric criterion but contains a material hallucination receives no credit. A panel of three judges grades each deliverable criterion-by-criterion. | 27 | 2.8% | |
| AA-AnalystAgent A data analysis benchmark, testing AI agents on their ability to work with spreadsheets and documents to answer quantitative questions a Business Analyst or Data Analyst would face day-to-day. | 26 | 40.0% | |
| EnterpriseOps-Gym-AA An independent implementation of ServiceNow's EnterpriseOps-Gym, an agentic benchmark testing whether LLM agents can complete stateful, multi-step enterprise workflows across eight business domains via live tool use, graded on the final state of the underlying databases. | 23 | 40.2% | |
| MATH-500 A 500-problem subset from the MATH dataset, featuring competition-level mathematics across six domains including algebra, geometry, and number theory. | 23 | 81.0% | |
| Code2Video-AA An implementation of HeyGen's Code2Video Bench, evaluating how well models create short product and brand videos from 114 design briefs using HyperFrames HTML. | 22 | 1025 | |
| Multimodal Knowledge Agent (MKA) An implementation of Google DeepMind's Multimodal Knowledge Agent (MKA) benchmark. It tests whether agents can read visually dense project files, such as architectural drawings, engineering schematics and clinical scans, and produce the deliverables a professional would. | 21 | 41.0% | |
| VIALS-AA VIALS-AA is an implementation of Handshake AI's VIALS benchmark. It tests whether language models can interpret scientific images and answer questions about the evidence they contain. | 21 | 45.3% | |
| AA-3D: Interactive 3D Experience A benchmark for how well AI models build interactive 3D experiences. Models work as agents in an offline sandbox via Stirrup to turn a written brief into a three.js scene, with Elo ratings derived from blind pairwise votes. | 20 | 995 | |
| CWE-Bench-AA An implementation of Collinear AI's CWE-bench, a defensive cybersecurity benchmark testing whether coding agents can audit a real open-source codebase, find a vulnerability in a described area of concern, and patch it without breaking legitimate behavior. | 19 | 55.8% | |
| CyberGym-E2E-AA An implementation of Berkeley RDI's CyberGym-E2E, a defensive cybersecurity benchmark that tests agents on finding real memory-safety vulnerabilities in C/C++ open-source projects, reproducing them with proof-of-concept inputs, and patching them without breaking the projects' tests. | 19 | 71.8% | |
| DeepsecBench-AA An implementation of Vercel's DeepsecBench, a defensive cybersecurity benchmark testing how well agents find vulnerabilities in real open-source application code. | 19 | 19.9% | |
| HealthBench OpenAI's HealthBench: 5,000 health conversations graded against physician-written rubrics, from emergency referrals to global health. | 15 | 58.5% | |
| HealthBench Professional OpenAI's HealthBench Professional: 525 clinician conversations covering consult, writing and research tasks, graded against physician-written rubrics. | 15 | 63.0% | |
| ITBench-AA An implementation of IBM's ITBench benchmark, testing AI agents on Kubernetes incident root-cause analysis from offline incident snapshots. The agent inspects alerts, events, traces, and topology and identifies the contributing-factor entities (deployments, pods, namespaces, network policies, etc.) responsible for the failure. | 14 | 39.2% | |
| APEX-Agents-AA An implementation of the APEX-Agents benchmark, testing AI agents on long-horizon, cross-application tasks in professional-services environments with realistic application tooling. | 9 | 27.4% |