GPT-5.4

Leaderboards

Where GPT-5.4 sits among the top 20 in each category.

By task

Scored from votes on each kind of prompt alone, on the same 1,000-point scale.

Test results

Every test GPT-5.4 has a published result on, out of 100%, with its rank among the models that took it.

• GPQA

  #12 / 256

  Graduate-level science questions written to be hard to look up.

  92.8%

  Self-reported

• Humanity's Last Exam

  #41 / 108

  Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions

  39.8%

  Self-reported · In the text score· In the vision score

• MMMU-Pro

  #9 / 73

  College-level questions that need both an image and text to answer.

  81.2%

  Self-reported · In the vision score

• BrowseComp

  #20 / 68

  Finding hard-to-locate facts by browsing the web.

  82.7%

  Self-reported · In the agents score

• SWE-Bench Pro

  #28 / 61

  Harder, longer software engineering tasks from real codebases.

  57.7%

  Self-reported · In the coding score

• Terminal-Bench 2.0

  #6 / 54

  Getting real work done in a command-line terminal.

  75.1%

  Self-reported

• DeepSWE 1.1

  #42 / 50

  DeepSWE 1.1 evaluates software engineering agents using the mini-swe-agent harness.

  52.0%

  Self-reported · In the coding score

• Toolathlon

  #15 / 42

  Tool Decathlon is a comprehensive benchmark for evaluating AI agents' ability to use multiple tools across diverse task categories. It measures proficiency in tool selection, sequencing, and execution across ten different tool-use scenarios.

  54.6%

  Self-reported · In the agents score

• LiveBench

  #3 / 38

  LiveBench is a challenging, contamination-limited LLM benchmark that addresses test set contamination by releasing new questions monthly based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses. It comprises tasks across math, coding, reasoning, language, instruction following, and data analysis with verifiable, objective ground-truth answers.

  80.3%

  Self-reported · In the text score

• MCP Atlas

  #27 / 38

  MCP Atlas is a benchmark for evaluating AI models on scaled tool use capabilities, measuring how well models can coordinate and utilize multiple tools across complex multi-step tasks.

  67.2%

  Self-reported

• Tau2 Telecom

  #3 / 36

  τ²-Bench telecom domain evaluates conversational agents in a dual-control environment modeled as a Dec-POMDP, where both agent and user use tools in shared telecommunications troubleshooting scenarios that test coordination and communication capabilities.

  98.9%

  Self-reported

• MRCR v2 (8-needle)

  #24 / 32

  MRCR v2 (8-needle) is a variant of the Multi-Round Coreference Resolution benchmark that includes 8 needle items to retrieve from long contexts. This tests models' ability to simultaneously track and reason about multiple pieces of information across extended conversations.

  22.8%

  Verified · In the text score

• OSWorld-Verified

  #15 / 28

  OSWorld-Verified is a verified subset of OSWorld, a scalable real computer environment for multimodal agents supporting task setup, execution-based evaluation, and interactive learning across Ubuntu, Windows, and macOS.

  75.0%

  Self-reported

• HealthBench Hard

  #8 / 24

  A challenging variation of HealthBench that evaluates large language models' performance and safety in healthcare through 5,000 multi-turn conversations with particularly rigorous evaluation criteria validated by 262 physicians from 60 countries

  26.0%

  Verified

• LongBench v2

  #15 / 24

  LongBench v2 is a benchmark designed to assess the ability of LLMs to handle long-context problems requiring deep understanding and reasoning across real-world multitasks. It consists of 503 challenging multiple-choice questions with contexts ranging from 8k to 2M words across six major task categories: single-document QA, multi-document QA, long in-context learning, long-dialogue history understanding, code repository understanding, and long structured data understanding.

  54.4%

  Verified

• FACTS Grounding

  #3 / 20

  A benchmark evaluating language models' ability to generate factually accurate and well-grounded responses based on long-form input context, comprising 1,719 examples with documents up to 32k tokens requiring detailed responses that are fully grounded in provided documents

  91.9%

  Verified

• OmniDocBench 1.5

  #8 / 20

  OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types. Lower Overall Edit Distance scores are better.

  89.1%

  Self-reported

• ARC-AGI v2

  #4 / 19

  ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks. It evaluates fluid intelligence via input-output grid pairs (1x1 to 30x30) using colored cells (0-9), requiring models to identify underlying transformation rules from demonstration examples and apply them to test cases. Designed to be easy for humans but challenging for AI, focusing on core cognitive abilities like spatial reasoning, pattern recognition, and compositional generalization.

  73.3%

  Self-reported · In the text score· In the vision score

• FrontierMath

  #4 / 17

  A benchmark of hundreds of original, exceptionally challenging mathematics problems crafted and vetted by expert mathematicians, covering most major branches of modern mathematics from number theory and real analysis to algebraic geometry and category theory.

  47.6%

  Self-reported

• FrontierSWE

  #10 / 16

  FrontierSWE measures whether an agent can complete open-ended technical projects at the scale of hours to tens of hours, spanning systems optimization, large-scale code construction, and applied ML research. Performance is reported as a dominance score, where higher is better.

  54.0%

  Self-reported

• Legal Agent Benchmark

  #11 / 15

  The Legal Agent Benchmark (LAB) is Harvey's open-source benchmark for evaluating AI agents on complex, long-horizon legal work. Tasks are scored under an all-pass standard against expert-curated rubrics, where a task passes only if every required rubric criterion (facts, conclusions, citations, structure, and analytical moves) passes.

  0.4%

  Self-reported

• Graphwalks BFS <128k

  #3 / 13

  A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length under 128k tokens, returning nodes reachable at specified depths.

  93.0%

  Self-reported

• HealthBench (hard)

  #2 / 13

  26.0%

  Unverified

• Graphwalks BFS >128k

  #9 / 12

  A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length over 128k tokens, testing long-context reasoning capabilities.

  21.4%

  Self-reported

• Graphwalks parents <128k

  #2 / 12

  A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length under 128k tokens, requiring understanding of graph structure and edge relationships.

  89.8%

  Self-reported

• ARC-AGI

  #3 / 11

  The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks. Each task consists of 2-5 demonstration pairs showing input grids transformed into output grids according to underlying rules, with test-takers required to infer these rules and apply them to novel test inputs. The benchmark uses colored grids (up to 30x30) with 10 discrete colors/symbols, designed to measure human-like general fluid intelligence and skill-acquisition efficiency with minimal prior knowledge.

  93.7%

  Self-reported

• mrcr-v2 (8needle_32768_65536)

  #5 / 10

  22.8%

  Unverified

• mrcr-v2 (8needle_4096_8192)

  #5 / 10

  55.1%

  Unverified

• mrcr-v2 (8needle_upto_128k)

  #5 / 9

  29.8%

  Unverified

• Finance Agent

  #6 / 8

  Finance Agent is a benchmark for evaluating AI models on agentic financial analysis tasks, testing their ability to process financial data, perform calculations, and generate accurate analyses across various financial domains.

  56.0%

  Self-reported

• facts-grounding (default)

  #3 / 8

  FACTS Grounding 1.0 — 860 public examples from Google DeepMind (CC-BY-4.0)

  91.9%

  Unverified

• Graphwalks parents >128k

  #4 / 7

  A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length over 128k tokens, testing long-context reasoning and graph structure understanding.

  32.4%

  Self-reported

• LongBench-V2 (short)

  #1 / 6

  55.6%

  Unverified

• LongBench-V2 (medium)

  #1 / 3

  54.4%

  Unverified