RankingEffort benchmarks

Gemini 3.5 Flash · High

Compare models

96 measured results · 4 effort settings · 70 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 3.5 Flash · High4 of 7 capabilities supported

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

Agentic
A45.3Supported
Hard reasoning
A44.3Supported
Coding
A35.2Supported
Human pref
Unknown
Knowledge
A62.1Supported
Multimodal
A60.5Preliminary
Long context
A47.2Preliminary

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 3.5 Flash · High: 38.0 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?

Datacurve · 15.2% direct comparison weight · independent of the model provider
ARC Prize · 15.2% direct comparison weight · independent of the model provider
Artificial Analysis · 15.1% direct comparison weight · independent of the model provider
Vals AI · 14.9% direct comparison weight · independent of the model provider
LiveBench · 14.7% direct comparison weight · independent of the model provider
Epoch AI · 13.2% direct comparison weight · independent of the model provider
Scale · 11.7% direct comparison weight · independent of the model provider

Which benchmark families carry the weight?

DeepSWE · 15.2% direct comparison weight
ARC-AGI · 15.2% direct comparison weight
enigma-eval · 11.7% direct comparison weight
epoch-game-puzzles · 4.6% direct comparison weight
simpleqa · 4.3% direct comparison weight
Other families (33) · 49.0% 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	82 → 84	Down 2	203 remaining / 241 lost	37.6 (-0.4 points)
ARC Prize	90 → 102	Down 12	444 remaining / 0 lost	38.2 (+0.2 points)
Datacurve	90 → 77	Up 13	444 remaining / 0 lost	39.9 (+1.9 points)
LiveCodeBench	90 → 90	No movement	444 remaining / 0 lost	38.0 (-0.0 points)
Scale	90 → 87	Up 3	444 remaining / 0 lost	41.0 (+3.0 points)
Terminal-Bench	90 → 92	Down 2	444 remaining / 0 lost	37.9 (-0.1 points)
Cognition	90 → 93	Down 3	444 remaining / 0 lost	37.8 (-0.2 points)
LiveBench	90 → 94	Down 4	443 remaining / 1 lost	36.9 (-1.1 points)
SWE-rebench	90 → 90	No movement	444 remaining / 0 lost	37.8 (-0.2 points)
Epoch AI	89 → 98	Down 9	431 remaining / 13 lost	37.2 (-0.8 points)
Vals AI	90 → 93	Down 3	443 remaining / 1 lost	36.7 (-1.3 points)
Provider reports · Anthropic	90 → 90	No movement	444 remaining / 0 lost	38.0 (+0.0 points)
Provider reports · Moonshot AI	90 → 91	Down 1	444 remaining / 0 lost	37.6 (-0.4 points)
Provider reports · Meta	90 → 90	No movement	444 remaining / 0 lost	38.0 (+0.0 points)
Provider reports · Mistral	90 → 90	No movement	444 remaining / 0 lost	38.0 (+0.0 points)
Provider reports · OpenAI	90 → 90	No movement	444 remaining / 0 lost	38.1 (+0.1 points)
All provider reports	90 → 91	Down 1	444 remaining / 0 lost	37.7 (-0.3 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	40.0	+2.0
frontiermath	39.2	+1.2
simpleqa	36.8	-1.2
enigma-eval	39.1	+1.1
mmmu-pro	39.0	+1.0
vals-finance-agent	37.0	-1.0
arc-agi	37.3	-0.7
swe-atlas-test-writing	38.6	+0.6
livebench-language	37.5	-0.5
livebench-instructions	37.5	-0.5
livebench-data	38.5	+0.5
vals-code-migration	38.4	+0.4
vibe-code	38.4	+0.4
livebench-reasoning	38.4	+0.4
enterprise-ops	37.7	-0.3
vals-tax-agent	38.3	+0.3
epoch-game-puzzles	38.3	+0.3
frontiercode	37.8	-0.2
apex-agents	37.8	-0.2
aime	37.8	-0.2
vals-legal-research	38.2	+0.2
swe-rebench	37.8	-0.2
swe-atlas-qna	38.2	+0.2
briefcase	38.2	+0.2
hle	37.8	-0.2
arc-agi-3	37.8	-0.2
analyst-agent	37.9	-0.1
critpt	38.1	+0.1
gpqa	37.9	-0.1
mmlu-pro	37.9	-0.1
terminal-bench	37.9	-0.1
livebench-math	38.1	+0.1
aa-lcr	38.1	+0.1
omniscience	38.1	+0.1
multi-swe-bench	38.1	+0.1
ifbench	37.9	-0.1
vals-excel-modeling	37.9	-0.1
gdpval	38.1	+0.1
gdp-pdf	38.1	+0.1
terminal-bench-science	38.1	+0.1
tau-bench	37.9	-0.1
harvey	38.1	+0.1
itbench	38.0	-0.0
automationbench	38.0	-0.0
livebench-coding	38.0	+0.0
swe-bench-pro	38.0	+0.0
math500	38.0	+0.0
gmmlu	38.0	+0.0
swe-atlas-refactoring	38.0	+0.0
livecodebench	38.0	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 104 observations, 85 contributing, 0 matched without graph weight and 19 excluded. Counts do not establish rank eligibility.

7 · Benchmark family or source protocol has not been reviewed
1 · 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
2 · Fewer than two distinct model identities remain in this evaluation
1 · Evaluator has not been reviewed for independent admission
4 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
1 · Composite index or publisher-specific internal evaluation lacks a stable general-capability comparison unit.
1 · Narrow specialist biological, medical or security evaluation; preserved as raw evidence outside general capability scoring.
2 · Supporting evidence outside the reviewed capability core

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.

Benchmark / source	Minimal	Low	Medium	High
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	Not reported	45% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	876.82 Elo Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	874.32 Elo Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
APEX-Agents-AAArtificial 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.	Not reported	Not reported	Not reported	47.05% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	61.33% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	74.33% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	73.33% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	42.08% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	1.43% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	10.86% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	13.14% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
EnterpriseOps-Gym-AAArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Overall task success rate across all eight enterprise domains (oracle tool mode) · Benchmark developed by ServiceNow Research · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-10-18.	Not reported	Not reported	Not reported	50.07% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	19.8% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	1,202.58 Elo Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	82.83% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	92.12% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	92.22% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	24.1% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	41.33% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	42.68% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
IFBenchArtificial 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.	47.28% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	74.56% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	76.33% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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	Not reported	40.35% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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	Not reported	18.33% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	80.12% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	83.87% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	84.28% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	0.67 Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	20.82 Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	21.18 Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	43.02% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	51.05% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	51.4% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	53.94% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
𝜏²-Bench TelecomArtificial 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.	58.77% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	95.61% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	95.32% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	32.16% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
Terminal-Bench HardArtificial 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.	46.21% Reported settings Gemini 3.5 Flash (Minimal); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-minimal	Not reported	39.39% Reported settings Gemini 3.5 Flash (Medium); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash-medium	40.91% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	78.65% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	Not reported	Not reported	Not reported	6.57% Reported settings Gemini 3.5 Flash (High); Model source: https://artificialanalysis.ai/models/gemini-3-5-flash
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.	8.89% Reported settings Gemini 3.5 Flash (Minimal); Source model ID: gemini-3-5-flash-minimal; 120 tasks	Not reported	Not reported	72.08% Reported settings Gemini 3.5 Flash (High); Source model ID: gemini-3-5-flash-high; 120 tasks
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	Not reported	Not reported	36.1% Reported settings gemini-3-5-flash; Reported effort: high; 113 unique tasks; 452 attempts; 4 runs

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

Gemini 3.5 Flash · High

Compare with

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context

Capability score evidence

Benchmark results