Grok 4.5

Leaderboards

Where Grok 4.5 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 Grok 4.5 has a published result on, out of 100%, with its rank among the models that took it.

• GPQA

  #10 / 256

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

  93.0%

  Self-reported

• SWE-Bench Pro

  #9 / 61

  Harder, longer software engineering tasks from real codebases.

  64.7%

  Self-reported · In the coding score

• DeepSWE 1.1

  #39 / 50

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

  54.0%

  Self-reported · In the coding score

• Terminal-Bench 2.1

  #18 / 46

  Terminal-Bench 2.1 is an updated release of the Terminal-Bench benchmark that tests AI agents' ability to operate a computer via the terminal. It evaluates how well models handle real-world, end-to-end tasks autonomously, including compiling code, training models, setting up servers, system administration, data science workflows, and security tasks.

  83.3%

  Self-reported

• Terminal-Bench 4.0

  #26 / 28

  Terminal-Bench 4.0 evaluates AI agents on real-world work in terminal and command-line containerized environments across 66 tasks, with emphasis on science-adjacent and frontier engineering problems. Relative to earlier Terminal-Bench releases, 4.0 increases timeouts and adaptively raises RAM/CPU on selected tasks to reduce harness and resource confounds.

  12.4%

  Self-reported · In the coding score· In the agents score

• FrontierCode 1.1

  #14 / 23

  FrontierCode 1.1 evaluates whether coding-agent changes are mergeable, using unit tests, maintainer-defined rubrics, and verifiers. Runs flagged for unfair internet use receive a zero score.

  42.4%

  Self-reported · In the coding score

• Artificial Analysis

  #7 / 13

  Artificial Analysis benchmark evaluates AI models across quality, speed, and pricing dimensions, providing a composite assessment of model capabilities for real-world usage.

  54.0%

  Self-reported

• DeepSWE

  #9 / 13

  DeepSWE is a software engineering agent benchmark evaluated with the mini-swe-agent harness, where each task is solved in an isolated container with no internet access. It measures an agent's ability to autonomously resolve real-world coding issues end to end.

  53.0%

  Self-reported

• Tau3 Banking

  #2 / 13

  τ³-Bench banking domain evaluates agentic models on multi-turn, tool-using customer-support scenarios in a simulated retail banking environment.

  33.0%

  Self-reported

• SWE-Marathon

  #9 / 10

  SWE-Marathon is an ultra-long-horizon software engineering benchmark covering tasks such as building compilers, optimizing kernels, and developing production-grade services. It measures whether agents can sustain quality across extremely long engineering trajectories.

  29.0%

  Self-reported

• OmniScience

  #2 / 6

  OmniScience is a broad scientific knowledge and reasoning benchmark that measures both answer accuracy and non-hallucination (calibrated abstention) across science domains.

  52.0%

  Self-reported

• AutomationBench-AA

  #1 / 5

  AutomationBench-AA is Artificial Analysis's independently run version of AutomationBench, covering 657 real-world SaaS workflow tasks across 40 simulated applications (e.g. Gmail, Slack, Salesforce, HubSpot). It scores the share of objectives an agent completes without violating business guardrails, using a private held-out task set.

  51.4%

  Self-reported

• AA-Omniscience Index

  #1 / 4

  AA-Omniscience Index is Artificial Analysis's knowledge-reliability metric. It rewards correct answers, penalizes hallucinations, and does not penalize abstention. Scores range from -100 to 100, where 0 means as many correct as incorrect answers.

  63.0%

  Self-reported

• DeepSWE 1.0

  #1 / 1

  DeepSWE 1.0 pass@1 benchmark as reported by Artificial Analysis using provider-specific harness runs.

  62.0%

  Unverified

• OmniScience (non-hallucination rate)

  #1 / 1

  OmniScience variant that reports the non-hallucination rate, defined as one minus the hallucination rate.

  46.0%

  Unverified