RankingEffort benchmarks

Gemini 3.8 Flash · High

Compare models

115 measured results · 4 effort settings · 76 benchmark/harness combinations

Model evidence and sources →

Change model

Model

Current configurationSources

Compare with

Choose a second model and its effort to see both profiles and benchmark differences below.

Compare withModel and configuration

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGemini 3.8 Flash · High4 of 7 capabilities supported

Gaps are unknown, not zero. Hollow points are preliminary.

Agentic
A50.8Supported
Hard reasoning
A63.1Supported
Coding
A68.4Supported
Human pref
Unknown
Knowledge
A76.2Supported
Multimodal
A68.4Preliminary
Long context
A64.0Preliminary

Each profile belongs to the named configuration. Capability scores use a shared panel for all axes, distinct from the overall ranking panel; they are not accuracy percentages or direct head-to-head wins.

Capability score evidence

Gemini 3.8 Flash · High: 71.8 overall comparison score. Meets the overall evidence requirements.

6 areas measured. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?

Terminal-Bench · 12.9% direct comparison weight · independent of the model provider
ARC Prize · 12.9% direct comparison weight · independent of the model provider
Epoch AI · 12.9% direct comparison weight · independent of the model provider
Vals AI · 12.8% direct comparison weight · independent of the model provider
LiveBench · 12.8% direct comparison weight · independent of the model provider
Artificial Analysis · 12.5% direct comparison weight · independent of the model provider
Cognition · 12.1% direct comparison weight · independent of the model provider
Datacurve · 11.2% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

Terminal-Bench · 15.0% direct comparison weight
FrontierCode · 12.1% direct comparison weight
DeepSWE · 11.2% direct comparison weight
ARC-AGI · 6.4% direct comparison weight
arc-agi-3 · 6.4% direct comparison weight
Other families (33) · 48.8% direct comparison weight

These shares describe direct comparison weight, not fractions of the final score or confidence. Source balancing does not give every family equal weight or remove correlations between overlapping sources.

DeepSWE v1.1 review: Epoch AI reported grading errors in at least 23 of 113 tasks. These published results remain included. Review their limitations in Epoch's review when interpreting this evidence.

How sensitive is this score?

The audit removes one source at a time and refits against the same overall reference panel. These published results use the matching dataset audited 2026-10-20.

Both ranks use the same surviving configurations for each removal. The cohort starts with 444 configurations supported in the original audit, including separate effort settings. These are diagnostic positions, not the default leaderboard ranks. Eligibility is not rechecked after removal.

What changes when a source is removed?
Removed source	Rank before → after	Rank movement	Cohort remaining / lost support	Score after (change)
Artificial Analysis	39 → 39	No movement	203 remaining / 241 lost	72.9 (+1.0 points)
ARC Prize	39 → 39	No movement	444 remaining / 0 lost	74.6 (+2.8 points)
Datacurve	39 → 40	Down 1	444 remaining / 0 lost	64.2 (-7.6 points)
LiveCodeBench	39 → 39	No movement	444 remaining / 0 lost	71.9 (+0.0 points)
Scale	39 → 38	Up 1	444 remaining / 0 lost	72.7 (+0.9 points)
Terminal-Bench	39 → 43	Down 4	444 remaining / 0 lost	66.1 (-5.8 points)
Cognition	39 → 26	Up 13	444 remaining / 0 lost	74.6 (+2.8 points)
LiveBench	39 → 33	Up 6	443 remaining / 1 lost	76.1 (+4.2 points)
SWE-rebench	39 → 39	No movement	444 remaining / 0 lost	71.3 (-0.5 points)
Epoch AI	39 → 39	No movement	431 remaining / 13 lost	72.9 (+1.0 points)
Vals AI	39 → 36	Up 3	443 remaining / 1 lost	71.8 (-0.1 points)
Provider reports · Anthropic	39 → 39	No movement	444 remaining / 0 lost	71.8 (-0.1 points)
Provider reports · Moonshot AI	39 → 39	No movement	444 remaining / 0 lost	71.5 (-0.3 points)
Provider reports · Meta	39 → 38	Up 1	444 remaining / 0 lost	71.8 (-0.1 points)
Provider reports · Mistral	39 → 39	No movement	444 remaining / 0 lost	71.8 (+0.0 points)
Provider reports · OpenAI	39 → 40	Down 1	444 remaining / 0 lost	71.8 (-0.1 points)
All provider reports	39 → 39	No movement	444 remaining / 0 lost	71.3 (-0.5 points)

Lost support means a configuration no longer has a comparison path to the complete reference panel. Its rank is unavailable. A surviving path alone does not establish leaderboard eligibility.

Overlapping sources may remain correlated even with source balancing. Leave-one-source-out movement is a sensitivity diagnostic, not a spread of source-specific ranks or a confidence measure. Scoring methodology · Dated configuration audit

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	64.2	-7.6
terminal-bench	66.0	-5.8
frontiercode	74.6	+2.8
arc-agi	73.8	+2.0
livebench-data	73.2	+1.3
frontiermath	73.1	+1.3
livebench-coding	73.0	+1.1
livebench-instructions	70.8	-1.1
gpqa	71.1	-0.8
vals-finance-agent	71.2	-0.7
vals-code-migration	72.5	+0.7
livebench-language	71.2	-0.6
mmmu-pro	72.4	+0.5
swe-rebench	71.3	-0.5
epoch-game-puzzles	71.4	-0.5
vals-legal-research	72.2	+0.4
vals-excel-modeling	71.5	-0.4
hle	72.2	+0.3
multi-swe-bench	71.5	-0.3
livebench-reasoning	71.6	-0.3
harvey	72.1	+0.3
simpleqa	71.6	-0.2
vibe-code	72.1	+0.2
automationbench	71.6	-0.2
tau-bench	71.6	-0.2
ifbench	71.6	-0.2
swe-atlas-qna	71.6	-0.2
itbench	71.6	-0.2
swe-atlas-test-writing	71.7	-0.2
omniscience	72.0	+0.2
gdp-pdf	72.0	+0.2
livebench-math	71.7	-0.2
analyst-agent	71.7	-0.1
critpt	72.0	+0.1
mmlu-pro	71.9	+0.1
aa-lcr	71.9	+0.1
arc-agi-3	71.8	-0.1
enigma-eval	71.9	+0.1
briefcase	71.9	+0.1
vals-tax-agent	71.9	+0.0
apex-agents	71.9	+0.0
terminal-bench-science	71.9	+0.0
swe-atlas-refactoring	71.8	-0.0
enterprise-ops	71.9	+0.0
aime	71.8	-0.0
livecodebench	71.9	+0.0
swe-bench-pro	71.9	+0.0
math500	71.9	+0.0
gmmlu	71.8	+0.0
gdpval	71.8	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 152 observations, 98 contributing, 0 matched without graph weight and 54 excluded. Counts do not establish rank eligibility.

6 · Benchmark family or source protocol has not been reviewed
3 · Original SciCode grading defects are under review; retained as supporting evidence, excluded from aggregate and capability scores. Corrected SciCode-Verified requires a separate protocol review. Sources: https://arxiv.org/abs/2608.04975 and https://artificialanalysis.ai/evaluations/scicode
6 · Evaluator has not been reviewed for independent admission
2 · Effort is not identified
13 · Supporting evidence outside the reviewed capability core
15 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
1 · No matched, identified effort unit; duplicates and unnamed comparators do not add evidence
1 · Card reports agentic result but linked methodology describes only static no-tools protocol.
1 · Mixed source task revisions and best-of-three versus provider reporting prevent a uniform common-core unit.
4 · Narrow specialist biological, medical or security evaluation; preserved as raw evidence outside general capability scoring.
2 · Composite index or publisher-specific internal evaluation lacks a stable general-capability comparison unit.

Inspect every observation and exclusion →

Benchmark results

These are published benchmark scores. Supported effort estimates also appear on the Capability leaderboard. Results from different harnesses occupy different rows. Missing cells stay unknown; no scores are copied between effort levels.

What these comparisons mean

Each source reports its own effort labels. Matching labels do not establish equal compute budgets or identical fallback behavior. Open Reported settings to inspect the exact model name, agent and task coverage. Multiple results at one level remain visible rather than selecting the best. Source-reported None or Non-reasoning labels are not treated as evidence that an API supports those settings.

These measurements use their own dated source collection. The leaderboard fits identified configurations jointly; it contains no pooled model ratings. Explicit thinking budgets are separate configurations. Read the methodology.

Download all effort benchmark results · 8,915 results across 259 models · Collected 2026-10-18

Benchmark / source	Low	Medium	High	Unspecified
AA-AnalystAgentArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks solved on all 5 of 5 attempts · 80 tasks, 5 attempts per task · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	46.25% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase v1.1 is an agentic knowledge work benchmark developed by Artificial Analysis. AA-Briefcase Elo is a combined metric that aggregates rubric pass rate, analytical quality Elo and presentation Elo · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	1,062.43 Elo Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	1,181.5 Elo Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	1,202.59 Elo Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
AA-LCR v1.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	80.67% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	84% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	81.33% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
AutomationBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of task objectives completed with no guardrail violations · Higher is better · Benchmark developed by Zapier · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	36.51% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	60.89% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	59.93% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
CritPtArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Argonne and UIUC, with contributions from 60+ researchers globally. CritPt is under review by its authors following external feedback · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	4% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	12.29% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	18.29% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
GDP.pdf: All-pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of attempts where every atomic criterion passed · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	18.6% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	22.8% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	21% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
GDPval-AA v2.1 LeaderboardArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Elo rating for performance on real-world work tasks · Anchored to DeepSeek V4.1 Flash (max) at 1600 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-10-18.	1,307.18 Elo Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	1,421.18 Elo Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	1,436.3 Elo Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
GPQA DiamondArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	92.02% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	93.54% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	95.25% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
Harvey LAB-AA v1.1: Hallucination-Gated All-Pass RateArtificial Analysis · Stirrup; Harvey private dataset v1.1.0; 120 tasks across 24 practice areas; up to 200 turns; internet disabled; GPT-6 Sol, Grok 4.7 and Claude Opus 5.5 rubric judges; GPT-6 Sol High two-pass hallucination check; hallucination-gated all-pass averaged across judges Benchmark details Average over tasks of the share of the judge panel finding every rubric criterion passed, with a task zeroed on any material hallucination · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	0% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
Humanity's Last ExamArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	37.07% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	42.12% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	47.82% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
ITBench-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Combined public and private Site Reliability Engineering (SRE) tasks · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	52.54% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
MLCR-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of tasks judged accurate, complete, and concise · Expert + compound (hard) sets · Benchmark developed by Wisedocs · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	21.67% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
MMMU-ProArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	84.51% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	84.22% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	85.61% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
AA-Omniscience IndexArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis Unit: index. Source collected 2026-10-18.	21.35 Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	28.58 Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	29.55 Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
AA-Omniscience AccuracyArtificial Analysis · 6000 questions; 1 attempt; correct-answer share; GPT-5.6 Luna medium grader Benchmark details AA-Omniscience Accuracy (higher is better) measures the proportion of correctly answered questions out of all questions, regardless of whether the model chooses to answer · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	52.22% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	52.98% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	54.6% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
SciCodeArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark curated by scientists across 16 disciplines. SciCode is under review following an independent audit of the dataset · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	54.98% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	55.09% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	56.6% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
𝜏³-BankingArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by Sierra Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	33.2% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	45.77% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	44.95% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
Terminal-Bench-Science 0.1Artificial Analysis · mini-swe-agent; Terminal-Bench-Science 0.1.0; 70 tasks; pass@1 averaged over 3 repeats Benchmark details Share of tasks passed (pass@1), averaged over three repeats · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	10% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
Terminal-Bench v2.1Artificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Benchmark developed by the Laude Institute and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	83.15% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	83.9% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	87.64% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
Terminal-Bench v4.0Artificial Analysis · mini-SWE-agent v2.4.6; 66 tasks; pass@1 averaged over 3 repeats; 500 steps; task timeout up to 8 hours Benchmark details Benchmark developed by the Laude Institute, Stanford, and open-source contributors · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	10.1% Reported settings Gemini 3.8 Flash (Low); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-low	19.7% Reported settings Gemini 3.8 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash-medium	19.7% Reported settings Gemini 3.8 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-8-flash	Not reported
ARC-AGI-2ARC Prize · v2 Semi-Private Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	77.5% Reported settings Gemini 3.8 Flash (Low); Source model ID: google-gemini-3-8-flash-low; 120 tasks	82.92% Reported settings Gemini 3.8 Flash (Medium); Source model ID: google-gemini-3-8-flash-medium; 120 tasks	89.17% Reported settings Gemini 3.8 Flash (High); Source model ID: google-gemini-3-8-flash-high; 120 tasks	Not reported
DeepSWE v1.1Datacurve · mini-swe-agent Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-10-18.	Not reported	71% Reported settings gemini-3-8-flash; Reported effort: medium; 113 unique tasks; 452 attempts; 4 runs	73.8% Reported settings gemini-3-8-flash; Reported effort: high; 113 unique tasks; 447 attempts; 4 runs	Not reported
ARC-AGI-3 · provider-adapterARC Prize · Provider Adapter Benchmark details ARC-AGI-3; Provider Adapter. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	15.07% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group google-gemini-3-8-flash; variant provider-adapter-low; public-demo task scores are excluded	24.2% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group google-gemini-3-8-flash; variant provider-adapter-medium; public-demo task scores are excluded	35% Reported settings ARC-AGI-3 Semi-Private verified summary; Provider Adapter harness; source group google-gemini-3-8-flash; variant provider-adapter-high; public-demo task scores are excluded	Not reported
ARC-AGI-3 · standardARC Prize · Standard Benchmark details ARC-AGI-3; Standard. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	5.99% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group google-gemini-3-8-flash; variant low; public-demo task scores are excluded	3.99% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group google-gemini-3-8-flash; variant medium; public-demo task scores are excluded	10.37% Reported settings ARC-AGI-3 Semi-Private verified summary; Standard harness; source group google-gemini-3-8-flash; variant high; public-demo task scores are excluded	Not reported
FrontierCode 1.1 · extended-chiselCognition · chisel Benchmark details FrontierCode 1.1; chisel. Open each result for its exact effort, task subset, metric and protocol details. Unit: percent. Source collected 2026-10-18.	Not reported	53.45% Reported settings FrontierCode 1.1; extended; 150 tasks; chisel; source effort medium; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	51.24% Reported settings FrontierCode 1.1; extended; 150 tasks; chisel; source effort high; weighted rubric score, not pass rate; unfair internet runs zeroed by evaluator; flag rate 0	Not reported

Gemini 3.8 Flash · High

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results