AgentBench / results
← All models

GPT-6 Luna (cloud)

1 tested weight format · Explore results, then inspect the answers.

Cloud reference · OpenAI via OpenRouter — openai/gpt-6-luna-20260922. Quality is scored with the same task bank and graders. All speed and desktop-memory measurements are excluded. Hosted precision and reasoning settings may differ from local GGUF runs.

Choose a tested configuration

Hosted model source

OpenAI via OpenRouter · Hosted UNDISCLOSED (provider-reported)

This result tested a cloud endpoint, not a downloadable GGUF.
Open hosted model →

12 unresolved holds

Setup, memory & provenance · Hosted UNDISCLOSED (provider-reported)

Hosted UNDISCLOSED (provider-reported) · Not recorded of weights · openrouter · context 16384 · GPU layers unrecorded

Parameter size unverified · unknown

Memory fit has not been verified for this configuration.

partial Hosted quality reference; not a local GGUF test. All speed and local memory measurements are excluded.

Recorded settings and thinking verification

Temperature provider default (not set) · sample index 42 · output limit 4096 · reasoning effort low

RAM labels describe the test computer, not this model’s memory requirement. Speed comparisons keep the 32 GB and 128 GB runs separate.

Runtime build: unrecorded · Cloud · OpenAI via OpenRouter · concurrency: 1

Thinking status: native reasoning · probe cloud-transport-v1 · scope Native reasoning condition; no verified off/on comparison. Provider-reported usage audited against 16K context. Sample indices are not provider seeds. Temperature omitted; independent provider-default samples..

Requested off does not prove an absence of reasoning. Saved checks do not establish unrecorded historical template or build details.

Overall across tasks

58–87 / 100

Nine categories, equally weighted · 6/9 eligible

Possible score range with unscored tasks receiving zero to full credit. Incomplete results remain unranked; this is not a confidence interval. Quality only; speed is separate. How the index works →

What does it do well?

Select a category to explore the answers. Bars show what happened on every planned task. Incorrect answers and failures to finish both earn zero; technical problems and untested tasks stay unscored.

Exact scores and task times by suite

Each suite has its own metric and fixed task bank. The balanced index weights categories equally; individual scores and evidence remain available below.

Test suiteReasoning · lowDetails
Everyday coding tasksTask pass rate
Pass rate
4/4 · 100%
Answers completed
4/4 · 100%
Correct / completed
4/4 · 100%
4 full credit12 technical / untested / held12 unscored; excluded from these rates.
4/16 scored · PreviewNo measured time
Larger practical tasksMean task credit
Average credit
—
Full-credit rate
0/0 · —
Answers completed
0/0 · —
Full credit / completed
0/0 · —
11 technical / untested / held11 unscored; excluded from these rates.
0/11 scored · PreviewNo measured time
Logic and number puzzlesMean puzzle credit
Average credit
88.2%
Full-credit rate
30/34 · 88%
Answers completed
34/34 · 100%
Full credit / completed
30/34 · 88%
30 full credit4 incorrect
34/34 scoredNo measured time
Published hard problemsMean answer credit
Average credit
88.8%
Full-credit rate
8/11 · 73%
Answers completed
11/11 · 100%
Full credit / completed
8/11 · 73%
8 full credit2 partial credit1 incorrect
11/11 scoredNo measured time
Following exact instructions · original campaignStrict prompt compliance
Pass rate
20/23 · 87%
Answers completed
23/23 · 100%
Correct / completed
20/23 · 87%
20 full credit3 incorrect
23/23 scoredNo measured time
Reading without guessingExact-match accuracy
Pass rate
13/16 · 81%
Answers completed
16/16 · 100%
Correct / completed
13/16 · 81%
13 full credit3 incorrect
16/16 scoredNo measured time
Calculating from a passageNumeric exact-match accuracy
Pass rate
4/7 · 57%
Answers completed
7/7 · 100%
Correct / completed
4/7 · 57%
4 full credit3 incorrect
7/7 scoredNo measured time
Reading research papersAnswer token F1
Average credit
32.0%
Full-credit rate
0/2 · 0%
Answers completed
2/2 · 100%
Full credit / completed
0/2 · 0%
2 partial credit6 technical / untested / held6 unscored; excluded from these rates.
2/8 scored · PreviewNo measured time
Resisting incorrect correctionsAll-branches pass rate
Pass rate
11/12 · 92%
Answers completed
12/12 · 100%
Correct / completed
11/12 · 92%
11 full credit1 incorrect
12/12 scoredNo measured time
Following exact instructions · 128 GB campaignStrict prompt complianceNot planned

Model analysis

This analysis covers this configuration across its recorded thinking conditions and test suites.

Use this hosted model at low reasoning effort for logic puzzles, selected difficult math problems, and source-based answers that must hold up when challenged. Those results were strong, while calculations from passages were uneven. Small-project coding is a promising preview, not an established strength; longer tasks with tools remain unassessed. It is a better-supported choice for focused question answering than for delegating project work.

Evidence for this judgment
Logic and number puzzles · Reasoning · low
Average score88.2/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes30 full credit · 0 partial · 4 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Published hard problems · Reasoning · low
Average score88.8/100 · Mean answer credit
Test coverage11 of 11 tasks scored
Task outcomes8 full credit · 2 partial · 1 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Resisting incorrect corrections · Reasoning · low
Tasks passed11 of 12 (91.7%)
Test coverage12 of 12 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Calculating from a passage · Reasoning · low
Tasks passed4 of 7 (57.1%)
Test coverage7 of 7 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Everyday coding tasks · Reasoning · low
Tasks passed4 of 4 (100.0%)
Test coverage4 of 16 tasks scored
Unresolved outcomes12 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Larger practical tasks · Reasoning · low
Average scoreNot measured
Test coverage0 of 11 tasks scored
Task outcomes0 full credit · 0 partial · 0 no credit
Unresolved outcomes11 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Source limitation
Some suite/condition results are incomplete or ineligible; their observed scores are previews.

Summary generated by GPT Astra · · Agent reviewed by Codex · publication authorized by user

About this summary
Analyst provider
openai
Generation route
Codex CLI · codex-cli 0.163.0 · ChatGPT account
Requested model
gpt-6-astra
Returned model
Not supplied; identity is recorded from the request only.
Exact provider snapshot
Exact provider snapshot not supplied.
Reasoning effort
high
Prompt version
model-summary-reader-v8
Generation time
2026-10-22T06:23:02.814475Z
Review time
2026-10-22T06:50:57.086342+00:00
Evidence reviewer
Codex (agent)
Publication authorization
User · 2026-10-22T05:55:43.625712+00:00 · Go ahead and complete the remaining work. Spin up agents: luna, terra, sol, astra, as needed to get it done. Narrative summaries: these are the astra generated summaries? Go ahead and get those completed as well. Once all of this is done, you may push the completed result.
Originating evidence edition
dataset-5a6f96ec56c3d240cc42fa23147c1495da1be71f3ef0722c0e5f8c4f5e126a83
Analysis input hash
6dc9e1a68b1853d36eed12b09c9e14bf858775a636b8a087faf0e32f9f20e360
Summary revision
summary-7cbcf4c79901df0636453e39f485b48c1a0c0e31b1549ae43da5ea7fb62fa90f

Generation and editorial authorship remain attached to this revision. Changing the default generator does not change this attribution.

Good at

A useful candidate for focused reasoning problems: it solved most logic and number puzzles and performed strongly on selected difficult math and reasoning questions. That supports bounded problem-solving, not a claim of broad mathematical mastery.

Evidence for this judgment
Logic and number puzzles · Reasoning · low
Average score88.2/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes30 full credit · 0 partial · 4 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Published hard problems · Reasoning · low
Average score88.8/100 · Mean answer credit
Test coverage11 of 11 tasks scored
Task outcomes8 full credit · 2 partial · 1 no credit
Explore these test results →

View this model’s answers, code and tool activity →

Answers passage questions well and generally follows strict wording and formatting rules. Strong correction-test results also support using it where answers must resist misleading challenges but change when the source evidence changes.

Evidence for this judgment
Resisting incorrect corrections · Reasoning · low
Tasks passed11 of 12 (91.7%)
Test coverage12 of 12 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Following exact instructions · Reasoning · low
Tasks passed20 of 23 (87.0%)
Test coverage23 of 23 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Reading without guessing · Reasoning · low
Tasks passed13 of 16 (81.3%)
Test coverage16 of 16 tasks scored
Explore these test results →

View this model’s answers, code and tool activity →

Struggles with

Calculations from passages need independent checking. Accuracy was mixed even on this small test, making it a shaky choice when counts and comparisons must be right.

Evidence for this judgment
Calculating from a passage · Reasoning · low
Tasks passed4 of 7 (57.1%)
Test coverage7 of 7 tasks scored
Incorrect completed answers3 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Research-paper answering remains a weak preview: the answer matching score was low on the small assessed portion, with most planned questions untested. Do not treat it as established research assistance.

Evidence for this judgment
Reading research papers · Reasoning · low
Average score32.0/100 · Answer token F1
Test coverage2 of 8 tasks scored
Task outcomes0 full credit · 2 partial · 0 no credit
Partially correct answers2 tasks
Untested tasks6 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Source limitation
Some suite/condition results are incomplete or ineligible; their observed scores are previews.

Worth knowing

Coding and research-paper results are incomplete previews; larger practical tasks remain unassessed. Unresolved and untested outcomes are not substantive wrong answers or confirmed failures to finish.

Evidence for this judgment

Source limitation
Some suite/condition results are incomplete or ineligible; their observed scores are previews.

Source limitation
Full and partial credit follow saved task checks. Incorrect answers and confirmed failures to finish within the test budget both earn zero, but are shown separately. Provider failures, ambiguous answer-channel failures and uncertain context guards remain unscored. A task blocked before execution is untested. Saved work can earn credit without a final chat message. Cloud timing is excluded. Authorized infrastructure retries count once per scheduled trial; original failures remain in history.

Everyday coding tasks · Reasoning · low
Tasks passed4 of 4 (100.0%)
Test coverage4 of 16 tasks scored
Unresolved outcomes12 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Reading research papers · Reasoning · low
Average score32.0/100 · Answer token F1
Test coverage2 of 8 tasks scored
Task outcomes0 full credit · 2 partial · 0 no credit
Untested tasks6 tasks
Explore these test results →

View this model’s answers, code and tool activity →

Larger practical tasks · Reasoning · low
Average scoreNot measured
Test coverage0 of 11 tasks scored
Task outcomes0 full credit · 0 partial · 0 no credit
Unresolved outcomes11 tasks
Explore these test results →

View this model’s answers, code and tool activity →

This hosted run establishes neither local speed nor memory needs. Native reasoning has no verified on/off comparison, so its added value is unknown; the output allowance includes reasoning rather than guaranteeing a separate thinking allowance.

Evidence for this judgment

Source limitation
Hosted provider/precision condition, not a local GGUF run. Cloud timing and memory are excluded from all local speed and memory comparisons. The hosted output allowance includes reasoning; it does not guarantee the separate local thinking cap. No verified thinking-on/off comparison was run.

Publication redactions limit reader inspection of some evidence.

Evidence for this judgment

Source limitation
Some selected evidence has recorded publication redactions.

How to interpret these assessments and their limitations →

How do the weight formats compare?

Everyday coding tasks · Task pass rate · requested / displayed reasoning on. Size is the weight download, not RAM. Different runtimes and placement can also affect results.

Hosted UNDISCLOSED (provider-reported) · Not recordedCloud · OpenAI via OpenRouter · openrouter · Unknown GPU layers · Recorded · some results unscored or qualifiedReasoning · low4/4 passed4/16 scored · 12 unresolved holds
Hosted UNDISCLOSED (provider-reported) · Not recordedCloud · OpenAI via OpenRouter · openrouter · Unknown GPU layers · Recorded · some results unscored or qualifiedReasoning · medium2/2 passed2/16 scored · 14 unresolved holds
Hosted UNDISCLOSED (provider-reported) · Not recordedCloud · OpenAI via OpenRouter · openrouter · Unknown GPU layers · Recorded · some results unscored or qualifiedReasoning · high2/2 passed2/16 scored · 14 unresolved holds
Hosted UNDISCLOSED (provider-reported) · Not recordedCloud · OpenAI via OpenRouter · openrouter · Unknown GPU layers · Recorded · some results unscored or qualifiedReasoning · xhigh2/2 passed2/16 scored · 14 unresolved holds
Hosted UNDISCLOSED (provider-reported) · Not recordedCloud · OpenAI via OpenRouter · openrouter · Unknown GPU layers · Recorded · some results unscored or qualifiedReasoning · maxNot scored0/16 scored · 16 unresolved holds

Qualified rows show the reason they are outside the rankings; they cannot establish superiority over a fully tested format. Open the setup to inspect context, runtime and controls.

Thinking tradeoff

No verified on/off comparison is available for this configuration and test suite. Requested modes alone do not establish a controlled thinking comparison.

Sampled repetitions

Repeated runs are a separate experiment and never add votes to the baseline score.

Reasoning · low · temperature

10/48 repeats scored · 10 full-credit repeats.

2/16 questions have every planned repeat scored. Of those, 2 earned full credit on every repeat.

Incomplete repetition coverage; unresolved questions are not classified as passing all repeats.

Recorded repetition conditions
  • Seed 303 · requested on · native reasoning
  • Seed 202 · requested on · native reasoning
  • Seed 101 · requested on · native reasoning

Everyday coding tasks: recorded outcomes

Find information in a small project, edit files and fix bugs.

4 full credit12 technical / untested / held

Each segment counts tasks, not credit points. Partial credit varies by task.

12 unresolved holds remain unscored.

Ranking unavailable: incomplete coverage.

Scoring and fixed question bank →

This model’s answers, code and tool activity

Select a question to read the original answer, tool calls and available code changes for this configuration and thinking setting.

1–16 of 16 matching questions · 16 in this version

  • Locate the Retry Limit

    The user asks where the project's retry-limit setting is defined and what value it holds. This exercises locating a named setting in code and reporting both its location and value.

    basic-T1 · primary
  • Find the Release Date

    The user asks for the release date mentioned in a project notes document. This exercises reading a specified file and extracting a particular fact rather than guessing from outside knowledge.

    basic-T2 · primary
  • Change the Retry Limit

    The user asks to change the project's retry limit to a specified value. This exercises finding the relevant setting and making a focused code edit rather than merely describing the change.

    basic-T3 · primary
  • Create an Exact One-Line File

    The user asks to create a text file containing exactly one supplied line. This exercises basic file creation and precise instruction following, including avoiding extra text in the file.

    basic-T4 · primary
  • Explain the Latest Commit

    The user asks what the most recent Git commit changed in the project. This exercises inspecting recorded changes and turning them into a clear explanation of what was modified.

    basic-T5 · primary
  • Count and Compare Python Files

    The user asks how many Python files are in the application folder and which is largest. This exercises identifying relevant files, counting them, and comparing their sizes.

    basic-T6 · primary
  • Add Readable Byte Sizes

    The user asks to add a function that converts byte counts into readable sizes and remove the related TODO comment. This exercises implementing a small utility and updating the surrounding code consistently.

    basic-T7 · primary
  • Collect TODO Comments

    The user asks to find every TODO comment in the Python code and append each to the project's task list on a separate line. This combines searching across files with collecting text into an existing document.

    basic-T8 · primary
  • Fix Code Behind a Failing Test

    The user asks to investigate a failing test and fix the application code without changing the test. This exercises diagnosing a bug while respecting an explicit limit on what may be edited.

    basic-H1 · primary
  • Rename a Function Throughout

    The user asks to rename a network function and update every place in the project that uses it. This exercises tracking references across files and making a coordinated change rather than editing only the definition.

    basic-H2 · primary
  • Find the Proxy Settings

    The user asks to find a proxy configuration file somewhere under the project's parent folder and report its host and port. This exercises searching beyond the project folder and reading configuration values.

    basic-H3 · primary
  • Locate an Error in a Log

    The user asks which error code appears in a long event log and on which line. This exercises finding a specific entry in a large file and reporting its precise location.

    basic-H4 · primary
  • Append a Final Task

    The user asks to append a supplied release-related task as the final line of an existing task list. This exercises preserving existing content while adding exact text in the requested position.

    basic-H5 · primary
  • Calculate Total Retry Waiting

    The user asks for the total time a network function spends sleeping when all attempts fail, worked out from the code. This exercises following retry logic and adding the delays along that failure path.

    basic-H6 · primary
  • Document Every Command-Line Option

    The user asks for a project readme listing every command-line option with its help text. This exercises discovering the program's available options and transferring that information into complete, source-grounded documentation.

    basic-H7 · primary
  • Undo a Commit by Editing

    The user asks to undo the latest Git commit's change by editing the file directly, without using Git to make changes. This exercises understanding a previous edit and reversing it under a method constraint.

    basic-H8 · primary
Filters change this list, not the fixed scoring denominator. Search currently covers question IDs, titles and available task explanations. Focus a control to load prompt/outcome search where this edition provides it.

Suggested comparisons

These are starting points for inspection, not claims that one setup is better. Runtime and placement differences remain visible.

No matching suggestion is available for this selection.

Tested weights

Hosted UNDISCLOSED (provider-reported) · Not recorded

OpenAI via OpenRouter · recorded source

Explore the answers

Choose a suite to inspect its exact available questions, references and model answers.