The leaders, at a glance
Agentic AI model ranking
Compare AI models and effort settings for agentic work using sourced task evaluations and visible evidence coverage.
Agentic leaders
Experimental comparison score · 0–100No supported leaders match this profile and filters. Try another profile, clear filters, or explore the results below.
For the selected profile and sources. Scores are experimental; small gaps may not be meaningful. How scoring works · Effort evidence collected
Explore all 109 sources Independent evaluations and lab reports
Distinct source pages with published results in the catalog. Multiple results and page sections count as one source.
- ai.meta.com
- Amazon Nova 2 technical report
- anthropic.com
- ARC Prize: ARC-AGI-2
- ARC Prize: ARC-AGI-3 · provider-adapter
- arcprize.org
- Artificial Analysis: AA-AnalystAgent
- Artificial Analysis: AA-Briefcase Elo
- Artificial Analysis: AA-LCR v1.1
- Artificial Analysis: AA-Omniscience Index
- Artificial Analysis: AIME 2025
- Artificial Analysis: APEX-Agents-AA
- Artificial Analysis: AutomationBench-AA
- Artificial Analysis: CritPt
- Artificial Analysis: EnterpriseOps-Gym-AA
- Artificial Analysis: GDP.pdf: All-pass
- Artificial Analysis: Global-MMLU-Lite
- Artificial Analysis: Harvey LAB-AA: Criterion Pass Rate
- Artificial Analysis: Humanity's Last Exam
- Artificial Analysis: IFBench
- Artificial Analysis: ITBench-AA
- Artificial Analysis: LiveCodeBench
- Artificial Analysis: MATH-500
- Artificial Analysis: MLCR-AA
- Artificial Analysis: MMLU-Pro
- Artificial Analysis: MMMU-Pro
- Artificial Analysis: SciCode
- Artificial Analysis: Terminal-Bench Hard
- Artificial Analysis: Terminal-Bench v2.1
- Artificial Analysis: Terminal-Bench v4.0
- Artificial Analysis: 𝜏²-Bench Telecom
- Artificial Analysis: 𝜏³-Banking
- artificialanalysis.ai
- artificialanalysis.ai
- artificialanalysis.ai
- artificialanalysis.ai
- artificialanalysis.ai
- artificialanalysis.ai
- Claude Fable 5.1 and Claude Mythos 5.1 System Card
- Cognition: FrontierCode 1.1 · extended-chisel
- Command A+ launch benchmarks
- Datacurve: DeepSWE v1.1
- deepmind.google
- deepseek-ai/DeepSeek-V4-Pro-0813
- deepswe.datacurve.ai
- Epoch AI: Epoch · Chess Puzzles · v1.0.0
- Epoch AI: Epoch · FrontierMath Tier 4 · v2.0.0
- Epoch AI: Epoch · FrontierMath Tiers 1–3 · v2.0.0
- Epoch AI: Epoch · GPQA diamond · v1.0.0
- Epoch AI: Epoch · Mystery Game Puzzles · v1.0.4 · Tool submission
- Epoch AI: Epoch · OTIS Mock AIME 2024-2025 · v1.0.0
- Epoch AI: Epoch · SimpleQA Verified · v1.0.0
- Gemini 3.8 Flash launch performance
- Gemini3.8Flash Model Card
- GPT-5.6: Frontier intelligence that scales with your ambition
- GPT-6 Astra System Card
- GPT-6 Astra: A new generation of intelligence
- Grok 4.6 launch evaluations
- huggingface.co
- huggingface.co
- huggingface.co
- Kimi K2.6 official model card
- kimi.ai
- LiveBench: LiveBench · AMPS Hard · 2026-06-25
- LiveCodeBench: LiveCodeBench
- livecodebench.github.io
- lmarena.ai
- MAI-Thinking-1 technical report
- MiniMax M3 model card benchmark figure
- Mistral Medium 3.5 model card performance charts
- moonshotai/Kimi-K3
- Muse Spark 1.1 evaluation report Figure44
- Muse Spark1.3 evaluation methodology
- nvidia/NVIDIA-Nemotron-3-Ultra-550B-A55B-BF16
- os-world.github.io
- qwen.ai
- Qwen/Qwen3.8-2.4T-A95B
- Scale: EnigmaEval · Scale official evaluation
- Scale: HLE · full · Scale official evaluation
- Scale: HLE · text only · Scale official evaluation
- Scale: SWE Atlas · QnA · claude code
- Scale: SWE Atlas · Refactoring · claude code
- Scale: SWE Atlas · Test Writing · claude code
- Scale: SWE-bench Pro public
- scale.com
- scale.com
- SWE-rebench: SWE-rebench · 2025-08 · tools
- swebench.com
- tbench.ai
- Terminal-Bench: Terminal-Bench 4.0 · claudecode
- Vals AI: Vals AI · Code Migration v1
- Vals AI: Vals AI · Excel Modeling Benchmark v1
- Vals AI: Vals AI · Finance Agent v2
- Vals AI: Vals AI · Harvey Legal Agent Benchmark v1
- Vals AI: Vals AI · Legal Research Bench v1
- Vals AI: Vals AI · MMLU-Pro v1
- Vals AI: Vals AI · Terminal-Bench 2.1
- Vals AI: Vals AI · Vibe Code Bench v1.1
- vals.ai
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- vulcanbench.com
- x.com
- zai-org/GLM-5.3
Customize ranking
Full ranking
127 entriesNo capabilities are weighted. Increase a weight to compare model performance.
| Rank | Model | Score | Capability coverage | analyst-agent | briefcase | aime | apex-agents | aa-lcr | automationbench | critpt | enterprise-ops | gdp-pdf | gdpval | gmmlu | gpqa | harvey | hle | ifbench | itbench | livecodebench | math500 | mmlu-pro | mmmu-pro | omniscience | scicode | tau-bench | terminal-bench | arc-agi | deepswe | swe-bench-pro | arc-agi-3 | frontiercode | enigma-eval | swe-atlas-qna | swe-atlas-refactoring | swe-atlas-test-writing | livebench-math | livebench-coding | livebench-language | livebench-data | multi-swe-bench | livebench-reasoning | livebench-instructions | swe-rebench | epoch-game-puzzles | frontiermath | simpleqa | vals-finance-agent | vals-code-migration | vals-legal-research | vals-excel-modeling | vibe-code | chartography | osworld | charxiv | swe-bench-multilingual | deepsearchqa | officeqa | mcp-atlas | toolathlon | mrcr | browsecomp |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 126 | Gemma 3 27B IT · Non-reasoningGoogle | 1.2 | Eligible · 100%22 published · 4 families · 1 selected capabilities | — | — | 20.0108/108 peersIndependent evidence | — | 12.6197/197 peersIndependent evidence | 3.189/89 peersIndependent evidence | 25.2200/200 peersIndependent evidence | — | 5.488/88 peersIndependent evidence | 0.9123/123 peersIndependent evidence | 30.763/63 peersIndependent evidence | 12.5212/212 peersIndependent evidence | — | 18.7208/208 peersIndependent evidence | 19.2178/178 peersIndependent evidence | — | 8.1112/112 peersIndependent evidence | 45.160/60 peersIndependent evidence | 22.4114/114 peersIndependent evidence | 18.2114/114 peersIndependent evidence | 16.5198/198 peersIndependent evidence | 5.291/91 peersIndependent evidence | 5.4202/202 peersIndependent evidence | 20.3201/201 peersIndependent evidence | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
| 127 | Phi-4-mini-instruct · Non-reasoningMicrosoft | 0.6 | Eligible · 100%15 published · 2 families · 1 selected capabilities | — | — | 4.9108/108 peersIndependent evidence | — | 19.7197/197 peersIndependent evidence | — | 25.5200/200 peersIndependent evidence | — | — | 0.0123/123 peersIndependent evidence | — | 6.5212/212 peersIndependent evidence | — | 19.3208/208 peersIndependent evidence | 3.8178/178 peersIndependent evidence | — | 5.8112/112 peersIndependent evidence | 11.560/60 peersIndependent evidence | 6.9114/114 peersIndependent evidence | — | 9.0198/198 peersIndependent evidence | — | 6.4171/171 peersIndependent evidence | 3.4201/201 peersIndependent evidence | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |
No models match these filters. Try another name or adjust the filters above.
About this ranking
This profile gives agentic work all the weight. Published agent-task evaluations contribute through matched benchmark families. Harnesses and reported settings remain attached to the evidence. These results describe the evaluated configurations; they do not establish how a model will perform in every workflow.
View the Capability ranking →Behind the ranking
273 entries with published results · 127 ranked · 0 provisional estimates.
Explore seven capabilities, the original results, and gaps in the evidence.
Cite this ranking
Include the model and effort setting, the selected profile, and the dated data downloads. A shared link uses the latest data; it does not freeze a ranking in time.
UnifyBench ranking Data updated: 12 Sept 2026 Mode: custom; evidence: documented; configurations: best Weights: v3;agentic:100,hard_reasoning:0,coding:0,human_pref:0,knowledge:0,multimodal:0,long_context:0 https://unifybench.ai/rankings/agentic?page=6
Catalog data · Effort measurements · Configuration scoring rules · Methodology
What does the score mean?
The adjusted comparison score estimates performance against the same reference panel from matched published comparisons. Capability pools matched comparisons across families and checks aggregate breadth and independent corroboration. Custom profiles require support in every selected capability. Missing results never count as losses.
Family details show observed win shares against available reference peers. Those descriptive values are supporting evidence, not adjusted capability scores. Provider reports and independent results remain labeled. Matching reported settings do not establish equal inference effort, and score differences do not establish statistical significance.
Read the full methodology →