AgentBench / results
← All models

Ministral 3 8B Instruct Q4_K_M

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

Choose a tested configuration

Get this tested model

Q4_K_M · 5.2 GB download

Links follow the selected quant and configuration.

Setup, memory & provenance · Q4_K_M

Q4_K_M · 5.2 GB of weights · vulkan · context 16384 · GPU layers 99

Medium · 5–under 20B · dense · 8.5B total

Memory fit has not been verified for this configuration.

partial

Recorded settings and thinking verification

Temperature 0 · seed 42 · output limit 4096 · reasoning effort unrecorded

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 · Local · test machine with 32 GB RAM · concurrency: 1

Thinking status: ignored switch · probe 2 · scope Saved profile-level probe; exact historical template/build scope is not fully recorded..

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

Overall across tasks

46.2 / 100

Nine categories, equally weighted · 9/9 eligible

Complete coverage for this setup. 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 suiteThinking offDetails
Everyday coding tasksTask pass rate
Pass rate
12/16 · 75%
Answers completed
16/16 · 100%
Correct / completed
12/16 · 75%
12 full credit4 incorrect
16/16 scored8.6s median task time · 16 measured
Larger practical tasksMean task credit
Average credit
45.5%
Full-credit rate
5/11 · 45%
Answers completed
11/11 · 100%
Full credit / completed
5/11 · 45%
5 full credit6 incorrect
11/11 scored29.7s median task time · 11 measured
Logic and number puzzlesMean puzzle credit
Average credit
26.6%
Full-credit rate
9/34 · 26%
Answers completed
34/34 · 100%
Full credit / completed
9/34 · 26%
9 full credit6 partial credit19 incorrect
34/34 scored19.1s median task time · 34 measured
Published hard problemsMean answer credit
Average credit
23.6%
Full-credit rate
2/11 · 18%
Answers completed
11/11 · 100%
Full credit / completed
2/11 · 18%
2 full credit3 partial credit6 incorrect
11/11 scored78.8s median task time · 11 measured
Following exact instructions · original campaignStrict prompt compliance
Pass rate
14/23 · 61%
Answers completed
23/23 · 100%
Correct / completed
14/23 · 61%
14 full credit9 incorrect
23/23 scored10.2s median task time · 23 measured
Reading without guessingExact-match accuracy
Pass rate
14/16 · 88%
Answers completed
16/16 · 100%
Correct / completed
14/16 · 88%
14 full credit2 incorrect
16/16 scored1.0s median task time · 16 measured
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 scored1.5s median task time · 7 measured
Reading research papersAnswer token F1
Average credit
31.1%
Full-credit rate
0/8 · 0%
Answers completed
8/8 · 100%
Full credit / completed
0/8 · 0%
7 partial credit1 incorrect
8/8 scored9.3s median task time · 8 measured
Resisting incorrect correctionsAll-branches pass rate
Pass rate
1/12 · 8%
Answers completed
12/12 · 100%
Correct / completed
1/12 · 8%
1 full credit11 incorrect
12/12 scored1.2s median task time · 12 measured
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 Instruct configuration for straightforward passage questions and small project edits you can check. Passage answers combine good accuracy with short waits, but small coding offers no clear speed advantage over successful alternatives on this test machine. Larger practical work is uneven, while handling challenged source-backed answers and difficult reasoning are substantial weaknesses. Treat it as a supervised helper rather than a dependable autonomous worker.

Evidence for this judgment
Reading without guessing · Thinking off
Tasks passed14 of 16 (87.5%)
Test coverage16 of 16 tasks scored
Tasks timed16 tasks
Typical time per task (median)1.0 seconds
Explore these test results →

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

Everyday coding tasks · Thinking off
Tasks passed12 of 16 (75.0%)
Test coverage16 of 16 tasks scored
Tasks timed16 tasks
Typical time per task (median)8.6 seconds
Explore these test results →

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

Larger practical tasks · Thinking off
Average score45.5/100 · Mean task credit
Test coverage11 of 11 tasks scored
Task outcomes5 full credit · 0 partial · 6 no credit
Explore these test results →

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

Resisting incorrect corrections · Thinking off
Tasks passed1 of 12 (8.3%)
Test coverage12 of 12 tasks scored
Explore these test results →

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

Published hard problems · Thinking off
Average score23.6/100 · Mean answer credit
Test coverage11 of 11 tasks scored
Task outcomes2 full credit · 3 partial · 6 no credit
Explore these test results →

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

Logic and number puzzles · Thinking off
Average score26.6/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes9 full credit · 6 partial · 19 no credit
Explore these test results →

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

Speed: Everyday coding tasks
Thinking off. Matching test-machine label and execution settings; spans model sizes. Timings include unsuccessful tasks, so read them alongside quality.

Speed and task success: Everyday coding tasksThinking off · Matching test-machine label and settings · Across model sizes

Shorter task time is better only when the result is useful. These times include failures, tools and grading. Token generation speed measures a different part of the work.

Compare timings and outcomes
ModelTypical task timeGeneration speedTasks passed
LFM2.5 1.2B Q4_01.4 seconds16 of 16 tasks timed296.5 tokens/s2 of 1616 of 16 scored
Phi-4 Mini Instruct 3.8B Q4_K_M3.9 seconds16 of 16 tasks timed102.8 tokens/s0 of 1616 of 16 scored
SmolLM3 3B Q4_K_M4.5 seconds16 of 16 tasks timed117.2 tokens/s3 of 1616 of 16 scored
Granite 4.0 Micro 3.4B Q4_05.6 seconds16 of 16 tasks timed107.8 tokens/s4 of 1616 of 16 scored
Qwen3.5 0.8B Q4_05.7 seconds16 of 16 tasks timed236.1 tokens/s6 of 1616 of 16 scored
Ling 3.0 Tiny Q4_05.8 seconds16 of 16 tasks timed122.4 tokens/s15 of 1616 of 16 scored
Qwen2.5 Coder 7B Instruct Q4_K_M6.8 seconds16 of 16 tasks timed66.5 tokens/s0 of 1616 of 16 scored
Ornith 1.5 9B Q4_K_M6.9 seconds16 of 16 tasks timed54.4 tokens/s15 of 1616 of 16 scored
Mellum 2.1 12B-A2.5B Thinking7.5 seconds16 of 16 tasks timed135.9 tokens/s14 of 1616 of 16 scored
Ministral 3 8B Instruct Q4_K_MThis model8.6 seconds16 of 16 tasks timed54.7 tokens/s12 of 1616 of 16 scored
Nemotron Nano 9B v2 Q4_08.9 seconds16 of 16 tasks timed52.9 tokens/s6 of 1616 of 16 scored
Granite 4.2 8B Q4_09.2 seconds16 of 16 tasks timed53.2 tokens/s14 of 1616 of 16 scored
Gemma 4 12B QAT Q4_013.3 seconds16 of 16 tasks timed35.2 tokens/s13 of 1616 of 16 scored

Results from this test setup, not a promise for your computer. Hardware, software, tokenizers and answer length can change the ordering. Historical runtime builds were not always captured. Comparison saved with the 2026-10-22 assessment.

Test machine: AMD Ryzen 5 5600X (6 cores / 12 threads) · AMD Radeon RX 6700 XT (12 GB VRAM) · 32 GB or 128 GB system RAM, never mixed in one timing group. Full test machine details

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
2026-10-22T05:53:47.604667Z
Review time
2026-10-22T05:56:50.673689+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
aef31a2bf11218c0871326d21f219ba1ae7c5de71b59372bcfaf9cacabb66de3
Summary revision
summary-b122ca819c3f7d15a073af9c986ce4fa90e7d4e73315012f6093d8074e8b2622

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

Good at

Answers straightforward passage questions well in a test mix that includes unanswerable questions, making it useful for extracting information from text.

Evidence for this judgment
Reading without guessing · Thinking off
Tasks passed14 of 16 (87.5%)
Test coverage16 of 16 tasks scored
Explore these test results →

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

Handles most bounded programming tasks in these tests, covering project-file searches, edits and bug fixes. Use it for small changes whose results you can check.

Evidence for this judgment
Everyday coding tasks · Thinking off
Tasks passed12 of 16 (75.0%)
Test coverage16 of 16 tasks scored
Explore these test results →

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

Struggles with

Struggles badly when source-backed answers must survive challenges or change with new evidence. Do not trust it to settle disputed facts without checking the source yourself.

Evidence for this judgment
Resisting incorrect corrections · Thinking off
Tasks passed1 of 12 (8.3%)
Test coverage12 of 12 tasks scored
Incorrect completed answers11 tasks
Explore these test results →

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

Larger practical tasks often missed the target, while puzzles and difficult math-and-reasoning problems scored poorly. Exact instruction-following also needs checking; useful small-project results do not carry over to these demands.

Evidence for this judgment
Larger practical tasks · Thinking off
Average score45.5/100 · Mean task credit
Test coverage11 of 11 tasks scored
Task outcomes5 full credit · 0 partial · 6 no credit
Incorrect completed answers6 tasks
Explore these test results →

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

Published hard problems · Thinking off
Average score23.6/100 · Mean answer credit
Test coverage11 of 11 tasks scored
Task outcomes2 full credit · 3 partial · 6 no credit
Explore these test results →

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

Logic and number puzzles · Thinking off
Average score26.6/100 · Mean puzzle credit
Test coverage34 of 34 tasks scored
Task outcomes9 full credit · 6 partial · 19 no credit
Explore these test results →

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

Following exact instructions · Thinking off
Tasks passed14 of 23 (60.9%)
Test coverage23 of 23 tasks scored
Explore these test results →

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

Compared with other models

Passage accuracy narrowly trails the leading tested medium-sized dense models; its task-time advantage over matching equally or more accurate alternatives is modest. Across sizes, several alternatives delivered better small-coding results sooner in these runs.

Evidence for this judgment
Reading without guessing · Thinking off
Tasks passed14 of 16 (87.5%)
Test coverage16 of 16 tasks scored
Tasks timed16 tasks
Typical time per task (median)1.0 seconds
Explore these test results →

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

Everyday coding tasks · Thinking off
Tasks passed12 of 16 (75.0%)
Test coverage16 of 16 tasks scored
Tasks timed16 tasks
Typical time per task (median)8.6 seconds
Explore these test results →

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

Reading without guessing
Thinking off. Medium · 5–under 20B · dense. Tied at rank 4 of 10 in this recorded comparison.

Speed: Reading without guessing
Thinking off. Matching test-machine label and execution settings; spans model sizes. Timings include unsuccessful tasks, so read them alongside quality.

Speed: Everyday coding tasks
Thinking off. Matching test-machine label and execution settings; spans model sizes. Timings include unsuccessful tasks, so read them alongside quality.

Reading without guessingThinking off · Medium · 5–under 20B · dense

Answer from a passage—or say that the passage does not contain the answer.

Exact-match accuracy · 0–100% · higher is better
16 of 16 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. =1 Gemma 4 12B11.9B total parameters93.8%
    1 task more passed
  2. =1 Gemma 4 12B QAT Q4_011.9B total parameters93.8%
    1 task more passed
  3. =1 Ornith 1.5 9B Q4_K_M9.2B total parameters93.8%
    1 task more passed
  4. =4 Ministral 3 8B Instruct Q4_K_M8.5B total parameters87.5%
    This model
  5. =4 Nemotron Nano 9B v2 Q4_08.9B total parameters87.5%
    Same score
  6. =6 Ministral 3 14B13.5B total parameters75.0%
    2 tasks fewer passed
  7. =6 Qwen2.5 Coder 7B Instruct Q4_K_M7.6B total parameters75.0%
    2 tasks fewer passed
  8. =6 Qwen3.5 9B9.0B total parameters75.0%
    2 tasks fewer passed
  9. #9 Gemma 4 E4B7.5B total parameters25.0%
    10 tasks fewer passed
  10. #10 Granite 4.2 8B Q4_08.8B total parameters12.5%
    12 tasks fewer passed
How this comparison works

Share of tasks that passed the recorded checks. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-22 assessment. Task-specific results.

Speed and task success: Reading without guessingThinking off · Matching test-machine label and settings · Across model sizes

Shorter task time is better only when the result is useful. These times include failures, tools and grading. Token generation speed measures a different part of the work.

Compare timings and outcomes
ModelTypical task timeGeneration speedTasks passed
LFM2.5 1.2B Q4_00.6 seconds16 of 16 tasks timed316.8 tokens/s0 of 1616 of 16 scored
Qwen3.5 0.8B Q4_00.6 seconds16 of 16 tasks timed243.6 tokens/s10 of 1616 of 16 scored
Phi-4 Mini Instruct 3.8B Q4_K_M0.7 seconds16 of 16 tasks timed108.6 tokens/s6 of 1616 of 16 scored
Ling 3.0 Tiny Q4_00.7 seconds16 of 16 tasks timed133.7 tokens/s10 of 1616 of 16 scored
Granite 4.0 Micro 3.4B Q4_00.7 seconds16 of 16 tasks timed113.9 tokens/s10 of 1616 of 16 scored
SmolLM3 3B Q4_K_M0.8 seconds16 of 16 tasks timed128.8 tokens/s1 of 1616 of 16 scored
Mellum 2.1 12B-A2.5B Thinking0.8 seconds16 of 16 tasks timed139.5 tokens/s13 of 1616 of 16 scored
Qwen2.5 Coder 7B Instruct Q4_K_M0.9 seconds16 of 16 tasks timed66.2 tokens/s12 of 1616 of 16 scored
Ministral 3 8B Instruct Q4_K_MThis model1.0 seconds16 of 16 tasks timed55.5 tokens/s14 of 1616 of 16 scored
Ornith 1.5 9B Q4_K_M1.1 seconds16 of 16 tasks timed54.0 tokens/s15 of 1616 of 16 scored
Nemotron Nano 9B v2 Q4_01.3 seconds16 of 16 tasks timed53.5 tokens/s14 of 1616 of 16 scored
Gemma 4 12B QAT Q4_01.4 seconds16 of 16 tasks timed37.3 tokens/s15 of 1616 of 16 scored
Granite 4.2 8B Q4_04.4 seconds16 of 16 tasks timed57.0 tokens/s2 of 1616 of 16 scored
Phi-4 Mini Reasoning 3.8B Q4_K_M4.9 seconds16 of 16 tasks timed103.5 tokens/s0 of 1616 of 16 scored

Results from this test setup, not a promise for your computer. Hardware, software, tokenizers and answer length can change the ordering. Historical runtime builds were not always captured. Comparison saved with the 2026-10-22 assessment.

Test machine: AMD Ryzen 5 5600X (6 cores / 12 threads) · AMD Radeon RX 6700 XT (12 GB VRAM) · 32 GB or 128 GB system RAM, never mixed in one timing group. Full test machine details

Choose Instruct for small coding: Ministral 3 8B Reasoning 2512 Q4_K_M matched its success but took longer. That separate checkpoint improved passage calculations and research-paper answer matching at substantially longer task times, yet did worse on larger practical tasks.

Evidence for this judgment
Everyday coding tasks · Thinking off
Tasks passed12 of 16 (75.0%)
Test coverage16 of 16 tasks scored
Tasks timed16 tasks
Typical time per task (median)8.6 seconds
Explore these test results →

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

Calculating from a passage · Thinking off
Tasks passed4 of 7 (57.1%)
Test coverage7 of 7 tasks scored
Tasks timed7 tasks
Typical time per task (median)1.5 seconds
Explore these test results →

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

Reading research papers · Thinking off
Average score31.1/100 · Answer token F1
Test coverage8 of 8 tasks scored
Task outcomes0 full credit · 7 partial · 1 no credit
Tasks timed8 tasks
Typical time per task (median)9.3 seconds
Explore these test results →

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

Larger practical tasks · Thinking off
Average score45.5/100 · Mean task credit
Test coverage11 of 11 tasks scored
Task outcomes5 full credit · 0 partial · 6 no credit
Explore these test results →

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

Everyday coding tasks
Thinking off. Ministral 3 8B · Instruct versus reasoning checkpoint. Tied at rank 1 of 2 in this recorded comparison.

Calculating from a passage
Thinking off. Ministral 3 8B · Instruct versus reasoning checkpoint. Rank 2 of 2 in this recorded comparison.

Reading research papers
Thinking off. Ministral 3 8B · Instruct versus reasoning checkpoint. Rank 2 of 2 in this recorded comparison.

Larger practical tasks
Thinking off. Ministral 3 8B · Instruct versus reasoning checkpoint. Rank 1 of 2 in this recorded comparison.

Everyday coding tasksThinking off · Ministral 3 8B · Instruct versus reasoning checkpoint

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

Task pass rate · 0–100% · higher is better
16 of 16 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. =1 Ministral 3 8B Instruct Q4_K_M8.5B total parameters75.0%
    This model
  2. =1 Ministral 3 8B Reasoning 2512 Q4_K_M8.5B total parameters75.0%
    Same score
How this comparison works

Share of tasks that passed the recorded checks. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-22 assessment. Task-specific results.

Calculating from a passageThinking off · Ministral 3 8B · Instruct versus reasoning checkpoint

Read a passage and count, compare or calculate using its facts.

Numeric exact-match accuracy · 0–100% · higher is better
7 of 7 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Ministral 3 8B Reasoning 2512 Q4_K_M8.5B total parameters100.0%
    3 tasks more passed
  2. #2 Ministral 3 8B Instruct Q4_K_M8.5B total parameters57.1%
    This model
How this comparison works

Share of tasks that passed the recorded checks. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-22 assessment. Task-specific results.

Reading research papersThinking off · Ministral 3 8B · Instruct versus reasoning checkpoint

Answer a question about a paper and identify the paragraphs supporting the answer.

Answer token F1 · 0–100 points · higher is better
8 of 8 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Ministral 3 8B Reasoning 2512 Q4_K_M8.5B total parameters52.7/100
    21.6 points higher
  2. #2 Ministral 3 8B Instruct Q4_K_M8.5B total parameters31.1/100
    This model
How this comparison works

Answer overlap with the reference answer; partial matches earn credit. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-22 assessment. Task-specific results.

Larger practical tasksThinking off · Ministral 3 8B · Instruct versus reasoning checkpoint

Build small apps, organize information and investigate files or web pages.

Mean task credit · 0–100 points · higher is better
11 of 11 tasks scored for each model.

Gaps are relative to this model. Treat small gaps cautiously.

  1. #1 Ministral 3 8B Instruct Q4_K_M8.5B total parameters45.5/100
    This model
  2. #2 Ministral 3 8B Reasoning 2512 Q4_K_M8.5B total parameters18.2/100
    27.3 points lower
How this comparison works

Average credit per task; partly correct results can earn partial credit. Family comparisons show the labeled Instruct and reasoning versions. Other peer comparisons use the same requested thinking setting, not necessarily each model’s best mode. Small differences on these short tests may not hold on other tasks. Peer membership varies by test coverage and compatible settings.

Comparison saved with the 2026-10-22 assessment. Task-specific results.

Worth knowing

The requested thinking-on condition is an ineligible preview without verified switch control. No within-checkpoint thinking recommendation is supported.

Evidence for this judgment

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

Source limitation
Some requested reasoning conditions lack verified control.

Publication redactions limit readers’ inspection of some evidence.

Evidence for this judgment

Source limitation
Some selected evidence has recorded publication redactions.

Tool-task results include selected retries after a role-format compatibility fix, not a fresh unbiased rerun. Instruct and Reasoning are separate checkpoints, not an on/off pair.

Evidence for this judgment

Source limitation
Earlier tool-task failures included a strict role-alternation rejection after context compaction. A scoped compatibility amendment moved the recap into the system message while preserving tool history. Retried technical failures are selected follow-ups, not a fresh unbiased pass-rate experiment. This tested Instruct checkpoint is not a thinking-off counterpart to the incoming Reasoning checkpoint.

How to interpret these assessments and their limitations →

One weight format is recorded for this model. Additional tested formats will appear here when available.

What changed in these results

Earlier tool-task failures included a strict role-alternation rejection after context compaction. A scoped compatibility amendment moved the recap into the system message while preserving tool history. Retried technical failures are selected follow-ups, not a fresh unbiased pass-rate experiment. This tested Instruct checkpoint is not a thinking-off counterpart to the incoming Reasoning checkpoint.

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.

Instruct vs Thinking / Reasoning

Two versions of the same model family, using separate GGUF checkpoints. Compare the extra task success with the extra time. This is a version comparison, not a switch within one file.

Everyday coding tasks

Scores and times describe this task bank. Quick failures are not a speed advantage. Different execution settings are labeled.

VersionTask resultTypical task timeWeight downloadExplore
Ministral 3 8B Reasoning 2512 Q4_K_MThinking / Reasoning version75.0%16/16 scored19.0 seconds16/16 timed5.2 GBAnswers and code

Compare

Ministral 3 8B Instruct Q4_K_MThinking off75.0%16/16 scored8.6 seconds16/16 timed5.2 GBAnswers and code

Sampled repetitions

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

Thinking off · temperature 0.6

48/48 repeats scored · 40 full-credit repeats.

16/16 questions have every planned repeat scored. Of those, 12 earned full credit on every repeat.

Recorded repetition conditions
  • Seed 101 · requested off · ignored switch
  • Seed 202 · requested off · ignored switch
  • Seed 303 · requested off · ignored switch

Everyday coding tasks: recorded outcomes

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

12 full credit4 incorrect

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

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.

Nemotron Nano 9B v2 Q4_0

Another checkpoint in the same total-parameter band and architecture category, with matching task controls.

Requested thinking off · Complete compared coverage

Compare these configurations →

Selection rule and evidence IDsPolicy comparisons-v3 · Template same-parameter-architecture-matching-controls
Facts: fact-eae080b73ed21e846eaa5d1d, fact-ec911e6561843ecc14d3482a
Matching candidate set: fact-ec911e6561843ecc14d3482a, fact-490482dd62af02e41d2fc20e, fact-6808af8c64a1e7875a14e7a6, fact-bef3eb6c6baa2d44e3aa0649, fact-2e2fe4760014b5b180de8620, fact-5c36b6f6bfeaa5ee466506f6, fact-6d911759dd69ca545ee8be5c, fact-1eaacebf5d4775d7901754f8, fact-f0e7d2286cd806072751d966, fact-9f9b21c2355c6bdb5b2dd393, fact-3f910f804bc145423c5f4359, fact-9b097653ded9dc087919acf0, fact-e2c619b31f8a88c1472ca60e, fact-e57043e6128b693616991807, fact-bb58072f1c77808045477c91, fact-81703cf46648d0b2de523fad, fact-200d3a80040bda993709f78f, fact-eacdd70113b5633f3f223c12

Tested weights

Q4_K_M · 5.2 GB

mistralai · recorded source

Ministral-3-8B-Instruct-2512-Q4_K_M.gguf · Exact GGUF file →Historical first-MiB hash only; not a full-file identity.

Explore the answers

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