RankingAgent's Last Exam

Agent's Last Exam

Data updated 22 Nov 2026

Agent's Last Exam is published here as its own result table. Scores use percent, and higher values rank better. The display harness is Google Gemini 4 Argon launch and evaluation methodology. Data updated 22 Nov 2026.

Bucket
Supporting evidence
Unit
percent
Direction
Higher is better
Version
—
Display harness
Google Gemini 4 Argon launch and evaluation methodology
Board
https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/

The available records have no admitted matched comparison in the capability core. Raw results remain available below.

Compare published benchmark results with category weights →

Models

Published configurations retain their source and harness labels. Missing results remain unknown.

1–3 of 3 entries

ModelScoreHarnessEvidenceSource-recorded date
Gemini 4 ArgonGoogle39.5%
Reported settings & source

Gemini API highest thinking settings; exact named effort not specified by Google. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent.

Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 15: Agent's Last Exam / Gemini 4 Argon; methodology Additional Details · reviewed 2026-09-30

Google Gemini 4 Argon launch and evaluation methodologylab self-report2026-09-30
Claude Opus 5.5Anthropic38.2%
Reported settings & source

Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent.

Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 15: Agent's Last Exam / Claude Opus 5.5; methodology Additional Details · reviewed 2026-09-30

Google Gemini 4 Argon launch and evaluation methodologylab self-report2026-09-30
GPT-6 AstraOpenAI34.2%
Reported settings & source

Comparator maximum available thinking/reasoning when reported, otherwise best available result; exact setting not identified by this grid. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent.

Provider-published launch claim, not independent measurement by Google for every row. Argon self-computed in default ALE-Claw harness with a five-hour window and safety filters. Flagged responses return empty strings; episode may continue. Astra and Opus from the official board. Binary pass rate. Fable absent. Methodology https://deepmind.google/models/evals-methodology/gemini-4-argon, SHA256 ff1df6bdeddc4c0f48840c09a8a7813b10ae7ad9e053892da68dabf64f09bff7. Preserve source rounding. Methodology says capabilities as of September 2026 and results as of October 2026; review date is source retrieval, not evaluation date.

Comparison limit: Provider-published claim. Independent source results are admitted separately with exact checkpoint, effort and compatible protocol; this grid does not establish a matched comparison.

Google Gemini 4 Argon launch and evaluation methodology · Launch grid row 15: Agent's Last Exam / GPT-6 Astra; methodology Additional Details · reviewed 2026-09-30

Google Gemini 4 Argon launch and evaluation methodologylab self-report2026-09-30