[LEADERBOARDS]
testing the limits of AI.
Benchmarks for frontier, agentic, and safety capabilities
Including benchmarks on agentic coding, frontier reasoning, and safety alignment.
From leading AI labs including OpenAI, Anthropic, Google, Meta, and open-source contributors.
DrugDiscoveryBench
DrugDiscoveryBench: 82 expert-curated tasks evaluating how reliably frontier coding agents perform the computational work of early-stage drug discovery.
GPT 6 Astra (mini-SWE-agent) max
NEW68.70±2.50
GPT 6 Astra (Codex) max
NEW65.40±1.90
Muse-Spark 1.3 (mini-SWE-agent) high
NEW62.20±2.10
SWE Atlas - Refactoring
Evaluating an agent's ability to restructure code while preserving behavior.
GPT 6 Astra (Codex) xHigh
NEW59.05±6.43
Fable-5.1 (Claude Code) xHigh
NEW56.67±6.52
Fable-5 (Claude Code) xHigh
54.76±6.76
SWE Atlas - Test Writing
Evaluating an agent’s ability to write production-grade tests
Fable-5.1 (Claude Code) xHigh*
NEW67.04±5.33
Opus 5 (Claude Code) xHigh
62.22±5.58
Fable-5 (Claude Code) xHigh*
55.60±5.80
SWE Atlas - Codebase QnA
Evaluating deep code comprehension and reasoning
Opus 5 (Claude Code) xHigh
63.17±5.01
Fable-5.1 (Claude Code) xHigh*
NEW59.95±4.85
GPT 6 Astra (Codex) xHigh*
NEW59.14±4.88
HiL-Bench (Human-in-Loop Benchmark)
Evaluates whether agents recognize information gaps and ask targeted clarifying questions.
Claude Fable 5.1
NEW61.50±6.47
Claude Opus 5
57.00±5.48
Claude Fable 5
56.33±5.50
MCP Atlas
Evaluating real-world tool use through the Model Context Protocol (MCP)
Muse Spark 1.1
88.10±1.95
Fable 5.1
NEW87.20±2.05
claude-opus-5 (xhigh)
85.80±2.10
SWE-Bench Pro (Public Dataset)
Evaluating long-horizon software engineering tasks in public open source repositories
Muse Spark 1.1*
NEW61.50±3.10
gpt-5.4 (xHigh)*
59.10±3.56
Muse Spark*
55.00±3.60
SWE-Bench Pro (Private Dataset)
Evaluating long-horizon software engineering tasks in commercial-grade private repositories
Muse Spark 1.1*
NEW51.50±5.50
claude-opus-4-6 (thinking)*
47.10±6.07
Muse Spark*
44.70±6.05
SciPredict
Forecasting scientific experiment outcomes
gemini-3-pro-preview
25.27±1.92
claude-opus-4-5-20251101
23.05±0.51
claude-opus-4-1-20250805
22.22±1.48
Humanity's Last Exam
Challenging LLMs at the frontier of human knowledge
54.80±1.94
46.50±2.00
46.44±1.96
Humanity's Last Exam (Text Only)
Challenging LLMs at the frontier of human knowledge
54.17±2.09
47.31±2.11
46.80±2.10
AudioMultiChallenge
Evaluating spoken dialogue systems in multi-turn interaction
gemini-3.8-flash (high)
NEW60.40±4.49
Inkling (Thinking)*
56.64±4.55
Inkling-small
54.87±4.57
AudioMultiChallenge - Audio Output
Evaluating spoken dialogue systems in multi-turn interaction
gpt-realtime-2 (xHigh)
48.45±4.59
tml-interaction-small
43.36±4.55
gpt-realtime-2
37.61±4.45
AudioMultiChallenge - Text Output
Evaluating spoken dialogue systems in multi-turn interaction
Inkling (Thinking)
NEW56.64±2.70
Inkling-small
NEW54.87±4.57
gemini-3-pro-preview (Thinking)
54.65±4.57
Professional Reasoning Benchmark - Finance
Evaluating Professional Reasoning in Finance
Muse Spark 1.3
NEW59.54±1.66
Muse Spark 1.1
55.01±0.14
claude-fable 5
53.86±0.14
Professional Reasoning Benchmark - Legal
Evaluating Professional Reasoning in Legal Practice
Muse Spark 1.3
NEW61.56±1.72
Muse Spark 1.1
57.05±0.22
claude-fable 5
52.56±0.54
Remote Labor Index (RLI)
Evaluating AI agents ability to perform real-world, economically valuable remote work
GPT 6 Astra
NEW20.83
Fable 5.1
NEW17.92
Fable 5
15.80
PropensityBench
Simulating real-world pressure to choose between safe or harmful behavior
o3-2025-04-16
10.50±0.60
claude-sonnet-4-20250514
12.20±0.20
o4-mini-2025-04-16
15.80±0.40
VisualToolBench (VTB)
Evaluating how LLMs can dynamically interact with and reason about visual information
Muse Spark 1.1
NEW44.77±2.82
gpt-5.4-2026-03-05 (reasoning effort = high)
29.17±0.13
gemini-3.1-pro-preview
28.97±0.91
MultiNRC
Multilingual Native Reasoning Evaluation Benchmark for LLMs
Muse Spark 1.1
NEW65.59±2.87
65.20±1.24
64.74±2.88
MultiChallenge
Assessing models across diverse, interdisciplinary challenges
Muse Spark
75.52±4.05
Muse Spark 1.1
NEW75.30±0.60
gemini-3.1-pro-preview
71.37±1.74
Fortress
Frontier Risk Evaluation for National Security and Public Safety
8.24±1.93
9.63±2.11
12.40±1.48
MASK
Evaluate model honesty when pressured to lie
96.28±0.41
96.13±0.57
Claude Sonnet 4 (Thinking)
95.33±2.29
EnigmaEval
Evaluating model performance on complex, multi-step reasoning tasks
claude-fable-5-high
39.28±2.80
gpt-5.6-sol-high
37.12±2.80
gemini-3.1-pro-preview-high
36.78±2.71
VISTA
Vision-Language Understanding benchmark for multimodal models
Gemini 2.5 Pro Experimental (March 2025)
54.65±1.46
gemini-2.5-pro-preview-06-05
54.63±0.55
gpt-5.4-pro-2026-03-05
53.89±2.02
TutorBench
Evaluating model performance on common tutoring tasks for high school and AP-level subjects
Muse Spark
68.55±0.95
gpt-5.4-pro-2026-03-05
56.62±1.02
gemini-2.5-pro-preview-06-05
55.65±1.11
Frontier AI Model Evaluations & Benchmarks
We conduct high-complexity evaluations to expose model failures, prevent benchmark saturation, and push model capabilities--while continuously evaluating the latest frontier models.
Scaling with Human Expertise
Humans design complex evaluations and define precise criteria to assess models, while LLMs scale evaluations--ensuring efficiency and alignment with human judgment.
Robust Datasets for Reliable AI Benchmarks
Our leaderboards are built on carefully curated evaluation sets, combining private datasets to prevent overfitting and open-source datasets for broad benchmarking and comparability.
[EVALUATE YOUR MODEL]
If you'd like to add your model to this leaderboard or a future version, please contact [email protected]. To ensure leaderboard integrity, we require that models can only be featured the FIRST TIME when an organization encounters the prompts.