# GPT-5.5

OpenAI · Proprietary · rank 24 · bench-align-v5

> Every figure below is reproduced as its upstream source published it: nothing is modelled, estimated, interpolated or converted. `Unavailable` means no source published the value — it is never a zero. Each value carries its evidence state and the date it was observed.

Page: https://modelscale.dev/models/gpt-5-5  
JSON: https://modelscale.dev/api/model/gpt-5-5

## Facts

| Field | Value | Evidence |
| --- | --- | --- |
| Canonical id | `gpt-5-5` | — |
| Overall score | 69 | Observed 2026-11-21 · source benchlm:models |
| Verified score (BenchLM): a separate BenchLM figure, not the overall score | 72 | Observed 2026-11-21 · source benchlm:models |
| Context window | 1M tokens | Observed 2026-11-21 · source benchlm:models |
| Max output (models.dev) | 128,000 tokens | Observed 2026-11-21 · source modelsdev:catalog |
| Modalities (models.dev) | Input: text, image, PDF · Output: text | Observed 2026-11-21 · source modelsdev:catalog |
| Release date | 2026-04-23 | Observed 2026-11-21 · source benchlm:models |
| Access type | Proprietary | — |
| Blended $/1M (75% input / 25% output) | $11.25 | Derived from the input and output rates below |

## Capability evidence

Seven axes from the ranking source. An axis the source did not score is unavailable, not zero. Beside each score is BenchLM's separate verified score for that axis; the two are published side by side and are not interchangeable.

| Axis | Score | Evidence | Verified (BenchLM) | Verified evidence |
| --- | --- | --- | --- | --- |
| Agentic | 81.4 | Observed 2026-11-21 · source benchlm:models | 78.2 | Observed 2026-11-21 · source benchlm:models |
| Coding | 57.1 | Observed 2026-11-21 · source benchlm:models | 55.1 | Observed 2026-11-21 · source benchlm:models |
| Knowledge | 78.2 | Observed 2026-11-21 · source benchlm:models | 78.2 | Observed 2026-11-21 · source benchlm:models |
| Reasoning | 67.4 | Observed 2026-11-21 · source benchlm:models | 67.1 | Observed 2026-11-21 · source benchlm:models |
| Multimodal & Grounded | 71.5 | Observed 2026-11-21 · source benchlm:models | 71.5 | Observed 2026-11-21 · source benchlm:models |
| Instruction Following | 93 | Observed 2026-11-21 · source benchlm:models | 93 | Observed 2026-11-21 · source benchlm:models |
| Math | 84.1 | Observed 2026-11-21 · source benchlm:models | 84.1 | Observed 2026-11-21 · source benchlm:models |

## Runtime service evidence

Measured values with the date they were observed. Nothing is inferred from a sibling model or a provider claim. Regional or per-endpoint measurements appear only when the API supplies them; none are modelled.

| Measurement | Value | Observed | Last good | Evidence |
| --- | --- | --- | --- | --- |
| Time to first token | 37.05 s | 2026-11-21 | 2026-11-21 | Observed 2026-11-21 · source benchlm:speed |
| Throughput | 98 tok/s | 2026-11-21 | 2026-11-21 | Observed 2026-11-21 · source benchlm:speed |

## Endpoint and price matrix

Every published price component, including cache reads and writes.

| Component | USD | Evidence |
| --- | --- | --- |
| Input / 1M tokens (base rate) | $5.00 | Observed 2026-11-21 · source benchlm:pricing |
| Output / 1M tokens (base rate) | $30.00 | Observed 2026-11-21 · source benchlm:pricing |
| Cache read / 1M tokens (base rate) | $0.50 | Observed 2026-11-21 · source benchlm:pricing |
| Cache write / 1M tokens (base rate) | Unavailable | Unavailable · source openrouter:pricing |
| Blended / 1M (75% input / 25% output) | $11.25 | Derived — from the input and output rates above; it has no source record of its own |
| Cost per successful task (LiveBench) | Unavailable | Unavailable · source livebench:table |
| Batch input / 1M tokens (OpenRouter batch) (base rate) | $2.50 | Observed 2026-11-21 · source openrouter:pricing |
| Batch output / 1M tokens (OpenRouter batch) (base rate) | $15.00 | Observed 2026-11-21 · source openrouter:pricing |
| Batch cache read / 1M tokens (OpenRouter batch) (base rate) | $0.25 | Observed 2026-11-21 · source openrouter:pricing |

Batch API rate (asynchronous jobs), from OpenRouter's batch listing.

### Long-context price tiers (models.dev)

The standard input, output and cache rates above are the base rate. models.dev publishes higher rates for requests whose prompt exceeds the thresholds below. The batch rates have their own long-context tiers, from OpenRouter's batch listing, shown separately. The cost simulator prices a standard-rate request whose whole prompt is above a threshold entirely at the highest of these tiers it exceeds, and a batch request only at a batch tier; it never applies these tiers to batch rates. The cost example on this page uses the base rate: its 1,200-token prompt is below every threshold.

| Prompt size | USD per 1M tokens | Evidence |
| --- | --- | --- |
| Above 272K prompt tokens | $10.00 input / $45.00 output per 1M · cache read $1.00 | Observed 2026-11-21 · source modelsdev:catalog |

### Long-context batch price tiers (OpenRouter batch)

The batch rates above are the base batch rate. OpenRouter's batch listing, the same listing as those rates, publishes higher batch rates for requests whose total prompt is above the thresholds below. The cost simulator prices a batch request whose whole prompt is above a threshold entirely at the highest batch tier it exceeds.

| Prompt size | Batch USD per 1M tokens | Evidence |
| --- | --- | --- |
| Above 272K prompt tokens | $5.00 input / $22.50 output per 1M · cache read $0.50 | Observed 2026-11-21 · source openrouter:pricing |

**Pricing note.** OpenAI's current API pricing page lists gpt-5.5 at $5.00 input / $0.50 cached input / $30.00 output per million tokens for short-context requests. Batch and Flex are available at half the standard rate, and Priority is 2.5x the standard rate.

## Workload-aware monthly cost example

10 conversations per day × 8 messages × 22 active days, 1200 input and 400 output tokens per message, no cache. Derived here from the published rates above by this site's own calculator — not a figure any source published.

| Field | Value |
| --- | --- |
| Modelled monthly cost | $31.68 |
| Modelled tokens | 2.82M |

## Benchmark record

89 matched benchmark rows with their published value, unit, and provenance.

### Knowledge

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| GPQA (Graduate-Level Google-Proof Q&A) | 93.6 | 448 questions | Graduate level | [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://arxiv.org/abs/2311.12022) |
| GPQA-D (GPQA Diamond) | 93.6 | Graduate-level science questions | Graduate level | [GPQA: A Graduate-Level Google-Proof Q&A Benchmark](https://arxiv.org/abs/2311.12022) |
| HLE (Humanity's Last Exam) | 52.2 | Expert-level questions | Frontier expert level | [Humanity's Last Exam](https://lastexam.ai/) |
| Artificial Analysis Intelligence Index | 38.4 | Cross-benchmark intelligence index | Display-only external reference | [Artificial Analysis](https://artificialanalysis.ai/) |
| AA-GPQA Diamond (Artificial Analysis GPQA Diamond) | 93.5 | Graduate-level science questions | Graduate-level science reasoning | [Artificial Analysis GPQA Diamond Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/gpqa-diamond) |
| AA-HLE (Artificial Analysis Humanity's Last Exam) | 45.8 | Expert-level questions | Frontier expert reasoning | [Artificial Analysis Humanity's Last Exam Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/hle) |
| AA-Omniscience Index (Artificial Analysis Omniscience Index) | 20.5 | Knowledge questions | Broad factual knowledge | [AA-Omniscience: Knowledge and Hallucination Benchmark](https://artificialanalysis.ai/evaluations/omniscience) |
| AA-Omniscience Accuracy (Artificial Analysis Omniscience Accuracy) | 58.0 | Knowledge questions | Broad knowledge | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| AA-Omniscience Hallucination Rate (Artificial Analysis Omniscience Hallucination Rate) | 89.0 | Knowledge questions | Factuality | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| HLE w/o tools (Humanity's Last Exam without tools) | 41.4 | Expert-level questions | Frontier expert level | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| GPQA Diamond (Vals) (GPQA Diamond, Vals AI run) | 93.2 | Graduate-level science questions | Expert reasoning | [Vals AI GPQA Diamond, Vals AI run leaderboard](https://www.vals.ai/benchmarks/gpqa) |
| MMLU-Pro (Vals) (MMLU-Pro, Vals AI run) | 88.1 | Academic multiple-choice questions | Broad academic knowledge | [Vals AI MMLU-Pro, Vals AI run leaderboard](https://www.vals.ai/benchmarks/mmlu_pro) |

### Coding

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| Terminal-Bench 2.0 | 82.0 | Terminal-based software tasks | Professional software engineering | [Terminal-Bench 2.0](https://www.tbench.ai/) |
| SWE-bench Pro | 58.6 | 1,865 repository problems | Long-horizon professional engineering | [SWE-Bench Pro: Can AI Agents Solve Long-Horizon Software Engineering Tasks?](https://arxiv.org/abs/2509.16941) |
| FrontierCode 1.1 Main | 43.0 | 100 private Main tasks (150 in Extended) | Frontier coding-agent quality | [FrontierCode leaderboard](https://cognition.com/frontiercode) |
| Vibe Code Bench (Vibe Code Bench v1.1) | 69.85 | End-to-end web application builds | End-to-end software delivery | [Vibe Code Bench: Evaluating AI Models on End-to-End Web Application Development](https://www.vals.ai/benchmarks/vibe-code) |
| React Native Evals | 84.7 | React Native app implementation tasks | Production mobile app engineering | [React Native Evals](https://rn-evals.vercel.app/) |
| AA Coding Index (Artificial Analysis Coding Index) | 74.9 | Cross-benchmark coding index | Display-only external reference | [Artificial Analysis model leaderboards](https://artificialanalysis.ai/leaderboards/models) |
| AA-SciCode (Artificial Analysis SciCode) | 55.8 | Scientific coding subproblems | Scientific programming | [Artificial Analysis SciCode Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/scicode) |
| LiveCodeBench (Vals) (LiveCodeBench, Vals AI run) | 85.3 | Competitive programming problems (easy, medium, hard) | Frontier coding | [Vals AI LiveCodeBench, Vals AI run leaderboard](https://www.vals.ai/benchmarks/lcb) |
| SWE-bench (Vals) (SWE-bench, Vals AI run) | 82.6 | Real repository issues by human time bucket | Frontier coding agents | [Vals AI SWE-bench, Vals AI run leaderboard](https://www.vals.ai/benchmarks/swebench) |
| PostTrainBench v1.1 | 27.2 | Post-training Qwen3 1.7B and 4B, SmolLM3 3B, and Gemma3 4B | Frontier agent evaluation | [PostTrainBench v1.1 leaderboard and methodology](https://posttrainbench.com/?version=v1.1) |
| PostTrainBench v1 (PostTrainBench v1 (archived)) | 25.0 | Post-training four base models on seven benchmarks | Autonomous ML engineering | [PostTrainBench v1 leaderboard (archived)](https://posttrainbench.com/?version=v1) |
| VulcanBench Frontier v4 | 78.5 | 23 behavioural-reconstruction coding tasks | Professional multi-file software engineering | [VulcanBench Frontier v4 leaderboard](https://vulcanbench.com/leaderboard.html) |
| FrontierCode 2.0 | 23.6 | 132 private software-engineering tasks | Frontier coding-agent quality | [FrontierCode 2.0](https://cognition.com/blog/frontier-code-2.0) |
| PostTrainBench (PostTrainBench v1.2) | 23.9 | Post-training four base models on six benchmarks | Autonomous ML engineering | [PostTrainBench](https://posttrainbench.com/) |

### Mathematics

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| FrontierMath (legacy) (FrontierMath legacy aggregate) | 51.7 | Historical aggregate | Research-level mathematics | [FrontierMath: A Benchmark for Evaluating Advanced Mathematical Reasoning in AI](https://epoch.ai/frontiermath) |
| FrontierMath v2 (Tiers 1-3) (FrontierMath v2 Tiers 1-3) | 85.263 | 285 private advanced mathematics problems | From olympiad-plus to early research mathematics | [FrontierMath v2 benchmark hub](https://epoch.ai/benchmarks/frontiermath-tiers-1-3-v2) |
| FrontierMath v2 (Tier 4) (FrontierMath v2 Tier 4) | 72.500 | 41 private extreme-difficulty mathematics problems | Research-level mathematics requiring hours or days of expert work | [FrontierMath Tier 4 v2 leaderboard](https://epoch.ai/benchmarks/frontiermath-tier-4-v2?view=graph&tab=leaderboard) |

### Reasoning

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| ARC-AGI-1 (ARC-AGI-1 Semi-Private Evaluation) | 95.00 | Semi-private ARC-AGI-1 evaluation set | Abstract visual reasoning | [ARC Prize leaderboard](https://arcprize.org/) |
| MRCR v2 64K-128K (OpenAI MRCR v2 8-needle 64K-128K) | 83.1 | 8-needle retrieval tasks | Long-context reasoning | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| MRCR v2 128K-256K (OpenAI MRCR v2 8-needle 128K-256K) | 87.5 | 8-needle retrieval tasks | Very long-context reasoning | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| ARC-AGI-2 (Abstraction and Reasoning Corpus for AGI v2) | 85 | Visual pattern completion and abstract reasoning | Expert-level — hardest public reasoning benchmark | [ARC-AGI-2: A Harder General Intelligence Benchmark](https://arcprize.org/arc-agi/2/) |
| ARC-AGI-3 (Abstraction and Reasoning Corpus for AGI v3) | 0.43 | Interactive game-like tasks with hidden rules | Frontier agentic reasoning | [ARC-AGI-3: A New Challenge for Frontier Agentic Intelligence](https://arcprize.org/media/ARC_AGI_3_Technical_Report.pdf) |
| AA-LCR (Artificial Analysis Long Context Reasoning) | 84.3 | Long-context reasoning tasks | Long-context reasoning | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| CritPt (Critical Physics Tasks) | 27.1 | Research-level physics questions | Research-level physics reasoning | [CritPt Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/critpt) |
| BioBasicsBench (BioBasicsBench: everyday wet-lab science) | 35.4 | 100 routine wet-lab analysis tasks (88 multimodal) across five domains | Routine wet-lab data interpretation | [Measuring how AI agents handle everyday wet-lab science](https://openai.com/index/measuring-how-ai-agents-handle-wet-lab-science/) |

### Instruction Following

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AA-IFBench (Artificial Analysis IFBench) | 75.9 | Verifiable instruction constraints | Instruction precision | [Artificial Analysis IFBench Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/ifbench) |

### Agentic

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| AA EnterpriseOps-Gym (Artificial Analysis EnterpriseOps-Gym) | 46.6 | Enterprise operations workflows | Enterprise agent operations | [Artificial Analysis EnterpriseOps-Gym Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/enterprise-ops-gym-aa) |
| AA ITBench (Artificial Analysis ITBench-AA) | 45.8 | IT incident-response tasks | Enterprise IT operations | [Artificial Analysis ITBench-AA Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/itbench-aa) |
| AA Tau3 Banking (Artificial Analysis Tau3-Banking) | 39.0 | Banking tool-use workflows | Agentic banking workflows | [Artificial Analysis Tau3-Banking Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/tau3-banking) |
| Terminal-Bench 2.0 | 82 | Terminal-based software tasks | Professional software engineering | [Terminal-Bench 2.0](https://www.tbench.ai/) |
| BrowseComp | 84.4 | Research questions requiring browsing | Hard web research | [BrowseComp](https://openai.com/index/browsecomp/) |
| GDPval-AA | 1396 | Agentic real-world work tasks | Professional agentic workflows | [DeepSeek-V4 Technical Report](https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro/blob/main/DeepSeek_V4.pdf) |
| GDPval-AA (GDPval-AA normalized) | 42.5 | Economically valuable tasks | Professional agentic workflows | [Artificial Analysis model benchmarks](https://artificialanalysis.ai/models/grok-4-3) |
| AA Agentic Index (Artificial Analysis Agentic Index) | 37.3 | Cross-benchmark agentic index | Display-only external reference | [Artificial Analysis model leaderboards](https://artificialanalysis.ai/leaderboards/models) |
| AA-AnalystAgent (Artificial Analysis AnalystAgent) | 50.0 | Spreadsheet and document analysis questions | Business and data analysis | [AA-AnalystAgent Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/aa-analyst-agent) |
| APEX-Agents-AA | 37.7 | 452 professional-services agent tasks | Long-horizon workplace agent tasks | [APEX-Agents-AA Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/apex-agents-aa) |
| Gert Labs (Gert Labs Composite Game Benchmark) | 72.93 | Novel game environments | Agentic coding and decision-making | [Gert Labs rankings](https://gertlabs.com/rankings) |
| OSWorld-Verified | 78.7 | 369 real-world computer tasks (361 when eight Google Drive tasks are excluded) | Multi-step desktop and cross-application workflows | [OSWorld](https://os-world.github.io/) |
| OSWorld 2.0 | 13.0 | 108 long-horizon computer-use workflows | Long-horizon professional workflows | [OSWorld2.0: Benchmarking Computer Use Agents on Long-Horizon Real-World Tasks](https://arxiv.org/abs/2606.29537) |
| CyberGym | 81.8 | 1,507 vulnerability analysis instances | Real-world cybersecurity | [CyberGym: Evaluating AI Agents' Real-World Cybersecurity Capabilities at Scale](https://www.cybergym.io/) |
| ExploitGym | 13.4 | 898 exploitation tasks | Advanced cybersecurity exploitation | [ExploitGym: Can AI Agents Turn Security Vulnerabilities into Real Attacks?](https://arxiv.org/abs/2605.11086) |
| JobBench | 42.7 | 130 tasks across 35 occupations | Professional multi-source workflows | [JobBench: Aligning Agent Work With Human Will](https://arxiv.org/abs/2605.26329) |
| MCP Atlas | 75.3 | Tool-integrated agent tasks | Advanced tool use | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| Toolathlon | 55.6 | Multi-tool workflows | Advanced tool use | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| τ²-bench results (τ²-Bench Tool-Agent-User Evaluation) | 98 | Airline, retail, and telecom customer-service task sets | Dual-control customer-service workflows | [τ²-Bench: Evaluating Conversational Agents in a Dual-Control Environment](https://arxiv.org/abs/2506.07982) |
| ResearchClawBench | 17.0 | 40 tasks across 10 scientific domains | Scientific research re-discovery | [ResearchClawBench: A Benchmark for End-to-End Autonomous Scientific Research](https://arxiv.org/abs/2606.07591) |
| Terminal-Bench 2.1 (Vals) (Terminal-Bench 2.1, Vals AI run) | 76.4 | Difficult terminal tasks | Frontier agentic | [Vals AI Terminal-Bench 2.1, Vals AI run leaderboard](https://www.vals.ai/benchmarks/terminal-bench-2-1) |
| ApprenticeBench (ApprenticeBench: end-to-end computer use, continual learning, and long-horizon agency on a real accounts-payable job) | 20 | 100 vendor bills processed in sequence inside a simulated construction company | Long-horizon computer use with offline and online continual learning | [ApprenticeBench: a step change in AI's job readiness](https://neocognition.io/blog/apprentice-bench/) |
| MazeBench (MazeBench: visual spatial reasoning in a 3D open world) | 0 | 100 hidden gems across 256 rooms in one open world | Long-horizon spatial planning, with single puzzles that can take over one hundred moves | [MazeBench Results](https://mazebench.com/blog?post=maze-bench-results) |
| LifeVal (LifeVal: everyday life tasks) | 77.4 | 169 everyday life tasks (caregiving, health, education, household administration, personal finance, shopping, government services) | Agentic personal-assistant work | [LifeVal](https://openai.com/index/lifeval/) |

### Multimodal & Grounded

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| MMMU-Pro (Massive Multi-discipline Multimodal Understanding Pro) | 81.2 | Multimodal academic reasoning | Frontier multimodal | [MMMU-Pro: A More Robust Multi-discipline Multimodal Understanding Benchmark](https://arxiv.org/abs/2409.02813) |
| AA-MMMU-Pro (Artificial Analysis MMMU-Pro) | 79.9 | Multimodal academic reasoning | Frontier multimodal | [Artificial Analysis MMMU-Pro Benchmark Leaderboard](https://artificialanalysis.ai/evaluations/mmmu-pro) |
| Design Arena Website (Design Arena Website Elo) | 1245 | Website generation comparisons | Design and website generation | [OpenRouter Grok 4.3 benchmarks](https://openrouter.ai/x-ai/grok-4.3/benchmarks) |
| OfficeQA Pro | 54.1 | Document and spreadsheet tasks | Enterprise grounded reasoning | [OfficeQA Pro: An Enterprise Benchmark for End-to-End Grounded Reasoning](https://arxiv.org/abs/2603.08655) |
| MMMU-Pro w/ Python (MMMU-Pro with Python) | 83.2 | Multimodal academic reasoning | Frontier multimodal | [Introducing GPT-5.4 mini and nano](https://openai.com/index/introducing-gpt-5-4-mini-and-nano/) |
| ParseBench-1 (ParseBench-1: document parsing for AI agents (LlamaIndex leaderboard, first release)) | 66.42 | 2,078 human-verified enterprise pages from insurance, finance and government documents | No method is strong on all five dimensions; charts and visual grounding separate the field | [ParseBench: A Document Parsing Benchmark for AI Agents](https://arxiv.org/abs/2604.08538) |
| ExtractBench (LlamaIndex) (ExtractBench: schema-guided enterprise document extraction (LlamaIndex leaderboard)) | 93.35 | 370 documents, 4,869 pages, 8 business domains, 67 document types, each with a JSON Schema | Short documents are near solved; long record lists and grounding separate systems | [ExtractBench: A Benchmark for Schema-Guided Enterprise Document Extraction](https://arxiv.org/abs/2607.29677) |

### external

| Benchmark | Value | Tasks | Difficulty | Provenance |
| --- | --- | --- | --- | --- |
| SEC-Bench Pro | 45.8 | Security engineering tasks | Advanced cybersecurity | [Introducing GPT-5.6](https://openai.com/index/gpt-5-6/) |
| Atomic network attacks (Atomic Network Attack Simulation) | 100 | Atomic cyber tasks | Network attack simulation | [Assessing GPT-5.6 Sol](https://www.irregular.com/research/assessing-gpt-5.6-sol) |
| Atomic vulnerability research (Atomic Vulnerability Research and Exploitation) | 92 | Atomic cyber tasks | Vulnerability research and exploitation | [Assessing GPT-5.6 Sol](https://www.irregular.com/research/assessing-gpt-5.6-sol) |
| Atomic evasion (Atomic Evasion) | 54 | Atomic cyber tasks | Cybersecurity evasion | [Assessing GPT-5.6 Sol](https://www.irregular.com/research/assessing-gpt-5.6-sol) |
| Chess Puzzles (Chess Puzzles (Epoch AI)) | 63.0 | 100 programmatically generated chess positions with a single best move | Novel chess positions, single best move | [Chess Puzzles (Epoch AI)](https://epoch.ai/benchmarks/chess-puzzles) |
| Mystery Game Puzzles (Mystery Game Puzzles (Epoch AI)) | 56.0 | 100 single-best-move positions from an undisclosed game | Unfamiliar game positions, single best move | [Mystery Game Puzzles (Epoch AI)](https://epoch.ai/benchmarks/mystery-game-puzzles) |
| Furniture Assembly (Furniture Assembly Benchmark (Epoch AI)) | 44.2 | 60 photos of three IKEA builds, some with deliberate assembly mistakes | Spatial reasoning from technical diagrams to real photos | [Furniture Assembly Benchmark (Epoch AI)](https://epoch.ai/benchmarks/furniture-assembly) |
| EBR-bench (Earthborne Rangers (EBR-bench)) | 34.3 | Ten playthroughs of an Earthborne Rangers campaign segment per sample (eight learning, two scored) | Learning an unfamiliar game over multi-hour playthroughs | [Earthborne Rangers (EBR-bench), Epoch AI](https://epoch.ai/benchmarks/ebr-bench) |
| MirrorCode (MirrorCode (Epoch AI and METR)) | 23.3 | 30 reimplementation tasks (15 programs in two languages), three attempts each | Days-long software engineering without source code | [MirrorCode (Epoch AI)](https://epoch.ai/benchmarks/mirrorcode) |
| Clocktower Bench (Clocktower Bench (Blood on the Clocktower)) | 1482 | 386 rated head-to-head games of Blood on the Clocktower (10 to 12 players) | Social deduction, deception and multi-agent reasoning | [Clocktower Bench leaderboard](https://clocktowerbench.com/) |
| ExploitGym Hardened (ExploitGym Hardened (stronger exploit mitigations)) | 2.3 | ExploitGym tasks with stronger exploit mitigations | Exploitation against mitigations | [ExploitGym: Can AI Agents Turn Security Vulnerabilities into Real Attacks?](https://arxiv.org/abs/2605.11086) |
| DAYJOB | 1.4 | 130 long-horizon tasks: 80 finance and 50 healthcare | Long-horizon professional work | [BenchLM](https://benchlm.ai/benchmarks/dayjob) |
| DAYJOB Healthcare | 0.0 | 50 long-horizon healthcare tasks across five subfields | Long-horizon clinical and healthcare operations work | [BenchLM](https://benchlm.ai/benchmarks/dayjob-healthcare) |
| Code2Video | 1571 | Creative briefs rendered as motion-graphics video | Code-driven motion design | [BenchLM](https://benchlm.ai/benchmarks/code2video) |
| Joke Arena | 1502 | 100 joke prompts in 10 categories, rated by human votes | Humour writing | [BenchLM](https://benchlm.ai/benchmarks/joke-arena) |
| Game Arena Go | 996 | Head-to-head Go games on a 13 x 13 board | Strategic planning and rule-following | [BenchLM](https://benchlm.ai/benchmarks/game-arena-go) |
| BabaIsBench (BabaIsBench (Baba Is You intro levels, pass@1)) | 100 | 8 intro levels of Baba Is You, three attempts per level (Stage 0); 15 further levels run once (Stage 1) | Rule-rewriting puzzles that need planning over many moves | [Baba Is Solved by Fable 5 and GPT-5.6 Sol, but at what cost? (Quesma)](https://quesma.com/blog/baba-is-bench/) |
| NYT Connections (NYT Connections LLM Benchmark (extended version)) | 96.2 | 940 NYT Connections puzzles, each with up to four added trick words | Wordplay, word knowledge and decoys that fit almost-but-not-quite | [NYT Connections LLM Benchmark (Lech Mazur, GitHub)](https://github.com/lechmazur/nyt-connections) |
| StealthMark (StealthMark Intelligence Benchmarks (overall score)) | 36.5 | 8 original tasks: Newton's Mirror, Ground Truth, Latch, Foresight, Mona Lisa, Selfie, Beach and Raccoon | Chained physical reasoning, planning without tools and precise visual construction | [StealthMark: The AI Benchmark Explained (Stealth Models)](https://stealthmodels.com/stealthmark) |
| SkillsBench official leaderboard (BenchFlow) (SkillsBench v1.1 official leaderboard (BenchFlow)) | 67.3 | 87 tasks in 8 domains (SkillsBench v1.1) | Multi-step agent tasks | [SkillsBench leaderboard](https://www.skillsbench.ai/leaderboard) |

## Independent scores (Epoch AI)

A second, independent score source: Epoch's own figures in Epoch's units, one line per model version Epoch evaluated (effort setting as Epoch labels it); higher is better. Not part of this site's capability scores or rank.

| Benchmark | Variant | Score | Unit | Epoch date | Evidence |
| --- | --- | --- | --- | --- | --- |
| GPQA Diamond | low | 0.9066 | fraction (0–1) | 2026-05-05 | Last good 2026-11-20 · source epoch:benchmarks |
| GPQA Diamond | none | 0.7727 | fraction (0–1) | 2026-08-07 | Last good 2026-11-20 · source epoch:benchmarks |
| FrontierMath Tiers 1-3 (v2) | xhigh | 0.8526 | fraction (0–1) | 2026-06-11 | Last good 2026-11-20 · source epoch:benchmarks |
| FrontierMath Tier 4 (v2) | xhigh | 0.725 | fraction (0–1) | 2026-06-11 | Last good 2026-11-20 · source epoch:benchmarks |
| SimpleQA Verified | xhigh | 0.63 | fraction (0–1) | 2026-08-27 | Last good 2026-11-20 · source epoch:benchmarks |
| Terminal-Bench | unknown (Codex CLI (developer's agent; 5 agents listed)) | 0.8225 | fraction (0–1) | 2026-04-23 | Last good 2026-11-20 · source epoch:benchmarks |
| OSWorld 2.0 | xhigh (batch tool) | 0.13 | fraction (0–1) | — | Last good 2026-11-20 · source epoch:benchmarks |

Source: [Epoch AI, CC BY 4.0](https://epoch.ai/benchmarks) ([licence](https://creativecommons.org/licenses/by/4.0/)). Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from https://epoch.ai/benchmarks.

## Lifecycle and limitations log

Lifecycle events the source associates with this model.

No lifecycle event references this model. That is not evidence the model has no lifecycle plan — only that this source published none.

## What this profile does not claim

Values are reproduced exactly as their sources published them, in the units those sources declared; none are converted, interpolated, or averaged across providers. Any field marked unavailable was attempted and not returned. Last attempted fetch for this model's score: 2026-11-21 01:17 UTC.
