Model
GPT-5.4
OpenAI · closed weights
Knowledge & Reasoning
Knowledge & Reasoning| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| CL-bench | 27.9 | overall | N/A | Independent | low | epoch.ai |
| CL-bench Life | 21.7 | overall | N/A | Independent | low | epoch.ai |
| Colombian Legal Reliability Benchmark | 85.9 | accuracy | N/A | Independent | medium | arxiv.org |
| CritPt | 23.4 | accuracy | N/A | Independent | low | epoch.ai |
| CroissantMiner | 0.7 | accuracy | single-pass | Independent | medium | arxiv.org |
| EnigmaEval | 16.0 | accuracy | N/A | Independent | medium | labs.scale.com |
| GameCommBench | 0.6 | analytical_completeness | N/A | Independent | medium | arxiv.org |
| GeoFitBench-LLM | 78.2 | success_rate | N/A | Independent | medium | arxiv.org |
| GPQA Diamond | 93.3 | accuracy | N/A | Independent | high | epoch.ai |
| GPQA Diamond | 92.8 | accuracy | N/A | Self-reported | low | artificialanalysis.ai |
| HLE | 36.2 | accuracy | N/A | Independent | low | epoch.ai |
| LMArena | 1463.0 | elo | N/A | Self-reported | low | lmarena.ai |
| MOF-VERIFY | 62.8 | macro-f1 | N/A | Independent | medium | arxiv.org |
| SimpleQA Verified | 45.1 | accuracy | N/A | Independent | high | epoch.ai |
Math
MathCoding
Coding| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| Closed-loop evaluation benchmark for embedded coding agents | 93.3 | pass_rate | N/A | Independent | medium | arxiv.org |
| DeepSWE | 51.8 | pass@1 | mini-swe-agent | Independent | medium | deepswe.datacurve.ai |
| GSO | 31.4 | score_opt@1 | OpenHands | Independent | medium | gso-bench.github.io |
| ModeBench | 0.3 | pcmd | N/A | Independent | medium | arxiv.org |
| QTestBench | 0.8 | test_quality | N/A | Independent | medium | arxiv.org |
| SciCode | 56.6 | score | N/A | Independent | medium | artificialanalysis.ai |
| SWE-bench Verified | 76.9 | resolved_rate | N/A | Independent | high | epoch.ai |
| TaoD2C-Bench | 60.2 | accuracy | N/A | Independent | medium | arxiv.org |
| Terminal-Bench 2.0 | 75.1 | accuracy | N/A | Self-reported | low | tbench.ai |
Agentic / Tool Use
Agentic / Tool Use| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| BrowseComp | 82.7 | accuracy | N/A | Self-reported | low | openai.com |
| DeepResearch Bench | 35.1 | average_score | N/A | Independent | medium | evals.futuresearch.ai |
| FinSkillBench | 0.5 | score | N/A | Independent | medium | arxiv.org |
| KlinikeBench | 23.4 | strict_pass@1 | N/A | Independent | medium | arxiv.org |
| LiveMACEBench | 0.7 | ras | N/A | Independent | medium | arxiv.org |
| MisVisAgentBench | 35.7 | pra | N/A | Independent | medium | arxiv.org |
| OpenGameEval | 40.0 | pass_rate | N/A | Independent | medium | arxiv.org |
| OSWorld | 75.0 | success_rate | N/A | Self-reported | low | os-world.github.io |
| Terminal-Bench | 81.8 | accuracy | ForgeCode | Independent | medium | tbench.ai |
Multimodal
Multimodal| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| BrainTRACE | 40.4 | accuracy | N/A | Independent | medium | arxiv.org |
| FigCodeBench | 79.7 | mmos | N/A | Independent | medium | arxiv.org |
| MM-FinEval | 0.7 | ret_l5m_da_↑_↑ | N/A | Independent | medium | arxiv.org |
| MMMU-Pro | 81.2 | accuracy | N/A | Self-reported | low | artificialanalysis.ai |
| OccludeBench | 49.5 | accuracy | N/A | Independent | medium | arxiv.org |
| SpaceCast-Bench | 37.7 | accuracy | N/A | Independent | medium | arxiv.org |
| SpatialChain | 72.2 | accuracy | N/A | Independent | medium | arxiv.org |
| WOVEN | 83.8 | accuracy | N/A | Independent | medium | arxiv.org |
Safety & Alignment
Safety & Alignment| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| AgentPrivArena | 17.3 | leak_rate_(%,_↓) | agentprivarena | Independent | medium | arxiv.org |
| LPHB | 0.8 | kappa | N/A | Independent | medium | arxiv.org |
Abstract Reasoning
Abstract ReasoningMultilingual
MultilingualReproduced by the community
Loading community runs…