RankingEffort benchmarks

Gemini 3.1 Pro effort benchmarks

Compare measured Low, Medium, High and other settings. Every number keeps its source.

Model

Gemini 3.1 Pro

75 measured results · 6 reported effort labels · 74 benchmark/harness combinations. Model evidence and sources →

Capability score evidence

Gemini 3.1 Pro · High: 36.2 overall comparison score. Meets the overall evidence requirements.

Which sources carry the weight?

Moonshot AI · 20.0% direct comparison weight · provider report
Datacurve · 14.8% direct comparison weight · independent
Scale · 14.8% direct comparison weight · independent
Vals AI · 14.2% direct comparison weight · independent
Epoch AI · 13.7% direct comparison weight · independent
LiveBench · 13.6% direct comparison weight · independent
Terminal-Bench · 8.9% direct comparison weight · independent

These shares describe the comparisons entering the fit, not fractions of the final score.

How sensitive is this score?

Remove one evaluator or family and refit the graph against the same overall panel. Values use the matching dataset audited 2026-09-12. These are sensitivity checks, not confidence intervals; ranking eligibility may also change.

Removed evidence	Recomputed score	Change
Artificial Analysis	33.5	-2.7
ARC Prize	37.6	+1.4
Datacurve	39.6	+3.4
LiveCodeBench	36.2	-0.0
Scale	33.5	-2.7
Terminal-Bench	38.2	+2.0
Cognition	36.8	+0.6
LiveBench	35.3	-1.0
SWE-rebench	36.1	-0.1
Epoch AI	33.6	-2.7
Vals AI	37.3	+1.1
Provider reports · Anthropic	36.2	-0.0
Provider reports · Moonshot AI	36.2	+0.0
Provider reports · Meta	36.2	-0.0
Provider reports · Mistral	36.2	+0.0
All provider reports	36.2	+0.0

Every benchmark family

Removed evidence	Recomputed score	Change
deepswe	39.6	+3.4
terminal-bench	38.3	+2.1
mmmu-pro	34.6	-1.6
hle	35.0	-1.2
arc-agi	37.4	+1.2
simpleqa	35.6	-0.6
mmlu-pro	35.6	-0.6
frontiercode	36.8	+0.6
aime	36.8	+0.5
livebench-instructions	35.8	-0.5
multi-swe-bench	36.6	+0.4
livebench-coding	36.5	+0.3
frontiermath	36.5	+0.3
livebench-language	36.0	-0.3
harvey	36.5	+0.2
livebench-reasoning	36.4	+0.2
gpqa	36.1	-0.2
vibe-code	36.4	+0.2
tau-bench	36.1	-0.2
swe-atlas-test-writing	36.4	+0.1
livebench-data	36.1	-0.1
swe-rebench	36.1	-0.1
vals-finance-agent	36.4	+0.1
apex-agents	36.1	-0.1
vals-code-migration	36.3	+0.1
vals-excel-modeling	36.3	+0.1
ifbench	36.1	-0.1
enigma-eval	36.1	-0.1
vals-legal-research	36.3	+0.1
arc-agi-3	36.1	-0.1
briefcase	36.3	+0.1
itbench	36.3	+0.1
swe-atlas-refactoring	36.1	-0.1
automationbench	36.3	+0.1
aa-lcr	36.2	-0.1
omniscience	36.3	+0.1
gdp-pdf	36.3	+0.1
enterprise-ops	36.3	+0.0
livebench-math	36.2	-0.0
scicode	36.2	-0.0
critpt	36.3	+0.0
livecodebench	36.3	+0.0
swe-atlas-qna	36.3	+0.0
math500	36.3	+0.0
gdpval	36.2	+0.0
gmmlu	36.2	+0.0
swe-bench-pro	36.2	-0.0
analyst-agent	36.2	+0.0
epoch-game-puzzles	36.2	-0.0

What evidence is missing from the fit?

Across all collected settings for this model: 184 observations, 71 contributing, 0 matched without graph weight and 113 excluded. Counts do not establish rank eligibility.

1 · Benchmark family or source protocol has not been reviewed
1 · Fewer than two distinct model identities remain in this evaluation
1 · Run-specific qualification requires review: In the initial run, 53 samples did not complete. These were rerun until 51 of them completed. The remaining samples are scored as incorrect.
1 · Run-specific qualification requires review: In the initial run, no epochs completed for 2/45 samples. Despite several re-runs of these two samples, no epochs completed. All epochs for the other 43/45 samples were correct. The model's score was edited to be 43/45.
2 · Effort is not identified
1 · Run-specific qualification requires review: In the initial run, 3% of samples failed to complete. Inspect defaults to not scoring these at all, but we override this behavior when >2% of epochs fail. For each sample, we average the score of the epochs that did complete for that sample. If none completed, we assign the random baseline of 25%. Then we average these scores across all samples. In this case, the resulting difference is very small: 94.1% modified vs. 94.4% original.
12 · Cited comparator requires original evaluation provenance; not a new matched run.
48 · Supporting evidence outside the reviewed capability core
43 · No reviewed effort for this catalog observation; any separately collected effort measurement is counted separately
1 · MiniMax cites this comparator from the official Terminal-Bench leaderboard; a common protocol with its internal Terminus 2 run is not established. See MiniMax M3 benchmark figure, Terminal-bench 2.1 methodology.
2 · Legacy self-report lacks a reviewed comparison configuration

Inspect every observation and exclusion →

Capability profile

One configuration, seven capabilities

Comparison score · 0–100

AGemini 3.1 Pro · High4 of 7 capabilities supported

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

Agentic
A38.5Supported
Hard reasoning
A56.3Supported
Coding
A29.8Supported
Human pref
Unknown
Knowledge
A79.4Supported
Multimodal
Unknown
Long context
Unknown

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.

Compare this configuration →Explore the model evidence profile →

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	Low	Medium	High	Thinking	Reasoning	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	41.25% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
AA-Briefcase EloArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details AA-Briefcase 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	458.83 Elo Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	32.01% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	82% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	35.43% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	17.71% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	42.17% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
GDP.pdf: All-passArtificial 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	17.8% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
GDPval-AA v2 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 a human baseline of 1,000 · Higher is better · Evaluation results measured independently by Artificial Analysis Unit: elo. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	903.94 Elo Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
Global-MMLU-LiteArtificial 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	93.17% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	94.14% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
Harvey LAB-AA: Criterion Pass RateArtificial Analysis · Artificial Analysis published evaluation; see source for benchmark-specific settings Benchmark details Share of atomic pass/fail rubric criteria the deliverables satisfy, graded by a single LLM judge · Independently benchmarked by Artificial Analysis · Evaluation results measured independently by Artificial Analysis Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	58.85% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	47.03% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	77.14% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	30.33% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	15.56% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	82.43% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	31.88 Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	54.85% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
SciCodeArtificial 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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	58.68% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	95.61% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	21.44% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	Not reported
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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	53.79% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	73.78% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	Not reported	Not reported	4.04% Reported settings Gemini 3.1 Pro Preview; Reported effort: reasoning; Model source: https://artificialanalysis.ai/models/gemini-3-1-pro-preview; Mode from evaluator isReasoning metadata; effort budget not specified	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 2026-09-12.	Not reported	Not reported	11.7% Reported settings gemini-3-1-pro-preview; Reported effort: high; 113 unique tasks; 452 attempts; 4 runs	Not reported	Not reported	Not reported
SWE-bench Pro publicScale · mini-swe-agent starred rows Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	46.1% Reported settings gemini-3.1-pro (thinking)*; 731 tasks; source date 2026-04-08	Not reported	Not reported
Terminal-Bench 2.1Terminal-Bench · Gemini CLI Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	65.84% Reported settings Gemini 3.1 Pro; Reported effort: high; 445 trials; recorded 2026-02-19	Not reported	Not reported	Not reported
Terminal-Bench 2.1Terminal-Bench · Terminus 2 Benchmark details Original public board. Configuration and task coverage may differ; inspect source. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	65.62% Reported settings Gemini 3.1 Pro; Reported effort: high; 445 trials; recorded 2026-02-19	Not reported	Not reported	Not reported
EnigmaEval · Scale official evaluationScale · Scale official evaluation Benchmark details Scale EnigmaEval puzzle-hunt accuracy; source-reported reasoning effort. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	36.78% Reported settings Scale EnigmaEval puzzle-hunt accuracy; source-reported reasoning effort.; published row 2025-11-06T22:52:23.000Z; reported margin ±2.71 percentage points; No contamination annotation reported (not proof of absence)	Not reported	Not reported	Not reported
HLE · text only · Scale official evaluationScale · Scale official evaluation Benchmark details Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	47.31% Reported settings Scale text-only public HLE subset; shares the full HLE family budget; automatic answer judging.; published row 2026-04-10T15:51:36.000Z; reported margin ±2.11 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	Not reported	Not reported	Not reported
HLE · full · Scale official evaluationScale · Scale official evaluation Benchmark details Scale full public HLE; 2,500 questions; automatic answer judging; temperature 0 when configurable. Public-set contamination caveats retained per result. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	46.44% Reported settings Scale full public HLE; 2,500 questions; automatic answer judging; temperature 0 when configurable. Public-set contamination caveats retained per result.; published row 2026-04-10T15:51:06.000Z; reported margin ±1.96 percentage points; Potential contamination warning: This model was evaluated after the public release of HLE, allowing model builder access to the prompts and solutions.	Not reported	Not reported	Not reported
LiveBench · AMPS Hard · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	98% Reported settings LiveBench official task evaluation; livebench-2026-06-25-AMPS_Hard; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · code completion · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	78.26% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_completion; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · code generation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	74.65% Reported settings LiveBench official task evaluation; livebench-2026-06-25-code_generation; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · connections · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	100% Reported settings LiveBench official task evaluation; livebench-2026-06-25-connections; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · consecutive events · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	85.24% Reported settings LiveBench official task evaluation; livebench-2026-06-25-consecutive_events; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · integrals with game · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	78% Reported settings LiveBench official task evaluation; livebench-2026-06-25-integrals_with_game; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · javascript · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	59.09% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-javascript; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · logic with navigation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	72% Reported settings LiveBench official task evaluation; livebench-2026-06-25-logic_with_navigation; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · math comp · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	96.08% Reported settings LiveBench official task evaluation; livebench-2026-06-25-math_comp; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · olympiad · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	92.1% Reported settings LiveBench official task evaluation; livebench-2026-06-25-olympiad; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · paraphrase · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	79.87% Reported settings LiveBench official task evaluation; livebench-2026-06-25-paraphrase; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · plot unscrambling · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	74.13% Reported settings LiveBench official task evaluation; livebench-2026-06-25-plot_unscrambling; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · python · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	50% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-python; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · simplify · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	71.05% Reported settings LiveBench official task evaluation; livebench-2026-06-25-simplify; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · spatial · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	98% Reported settings LiveBench official task evaluation; livebench-2026-06-25-spatial; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · story generation · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	77.25% Reported settings LiveBench official task evaluation; livebench-2026-06-25-story_generation; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · summarize · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	88.23% Reported settings LiveBench official task evaluation; livebench-2026-06-25-summarize; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · tablejoin · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	52.35% Reported settings LiveBench official task evaluation; livebench-2026-06-25-tablejoin; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · tablereformat · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	98.04% Reported settings LiveBench official task evaluation; livebench-2026-06-25-tablereformat; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · theory of mind · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	80.77% Reported settings LiveBench official task evaluation; livebench-2026-06-25-theory_of_mind; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · typescript · 2026-06-25LiveBench · LiveBench Mini-SWE-Agent, 250-step limit Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	23.33% Reported settings LiveBench Mini-SWE-Agent, 250-step limit; livebench-2026-06-25-typescript; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · typos · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	82% Reported settings LiveBench official task evaluation; livebench-2026-06-25-typos; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
LiveBench · zebra puzzle · 2026-06-25LiveBench · LiveBench official task evaluation Benchmark details LiveBench 2026-06-25 task snapshot. Task scores stay separate and related tasks share their reviewed family budget. Explicit checkpoint and effort labels only. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	85.25% Reported settings LiveBench official task evaluation; livebench-2026-06-25-zebra_puzzle; gemini-3.1-pro-preview-high; checkpoint=gemini-3.1-pro	Not reported	Not reported	Not reported
Epoch · Chess Puzzles · v1.0.2Epoch AI · Epoch AI Inspect; task 1.0.2; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	Not reported	55% Reported settings Epoch AI Inspect; task 1.0.2; scorer model_extracted_exact_match; model=gemini-3.1-pro-preview; run=4QfM9qyeWGqNc8YTcQNV76; mean accuracy; https://epoch.ai/data/benchmarks.csv Run-specific qualification requires review: In the initial run, 53 samples did not complete. These were rerun until 51 of them completed. The remaining samples are scored as incorrect. Source label needs review
Epoch · Chess Puzzles · v1.1.6Epoch AI · Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	49% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gemini-3.1-pro-preview_high; run=4mzBGLow3HkS8BmAYCzGTR; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported
Epoch · FrontierMath Tier 4 · v2.0.0Epoch AI · Epoch AI Inspect; task 2.0.0; scorer verification_code Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. FrontierMath was developed with OpenAI funding and unequal access to part of the question set; this result is from Epoch's evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	Not reported	26.83% Reported settings Epoch AI Inspect; task 2.0.0; scorer verification_code; model=gemini-3.1-pro-preview; run=eXVWr3RBeSsWy5DAU8ed46; mean accuracy; https://epoch.ai/data/benchmarks.csv
Epoch · FrontierMath Tiers 1–3 · v2.0.0Epoch AI · Epoch AI Inspect; task 2.0.0; scorer verification_code Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. FrontierMath was developed with OpenAI funding and unequal access to part of the question set; this result is from Epoch's evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	Not reported	59.65% Reported settings Epoch AI Inspect; task 2.0.0; scorer verification_code; model=gemini-3.1-pro-preview; run=hVifTUdQZrFEYabgCRu8Qq; mean accuracy; https://epoch.ai/data/benchmarks.csv
Epoch · GPQA diamond · v1.0.11Epoch AI · Epoch AI Inspect; task 1.0.11; scorer choice Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	94.44% Reported settings Epoch AI Inspect; task 1.0.11; scorer choice; model=gemini-3.1-pro-preview_high; run=BgmrBFsbsaD8rNpebv49Da; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported
Epoch · GPQA diamond · v1.0.5Epoch AI · Epoch AI Inspect; task 1.0.5; scorer choice Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	Not reported	94.1% Reported settings Epoch AI Inspect; task 1.0.5; scorer choice; model=gemini-3.1-pro-preview; run=kR9F3oHBCnwQW4AiTPKwpd; mean accuracy; https://epoch.ai/data/benchmarks.csv Run-specific qualification requires review: In the initial run, 3% of samples failed to complete. Inspect defaults to not scoring these at all, but we override this behavior when >2% of epochs fail. For each sample, we average the score of the epochs that did complete for that sample. If none completed, we assign the random baseline of 25%. Then we average these scores across all samples. In this case, the resulting difference is very small: 94.1% modified vs. 94.4% original. Source label needs review
Epoch · Mystery Game Puzzles · v1.0.4 · Tool submissionEpoch AI · Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	34% Reported settings Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=gemini-3.1-pro-preview_high; run=g4bbCaiFDqw2YYNDAcmn2a; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported
Epoch · Mystery Game Puzzles · v1.0.4 · Exact moveEpoch AI · Epoch AI Inspect; task 1.0.4; scorer exact_match_next_move_scorer Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	29% Reported settings Epoch AI Inspect; task 1.0.4; scorer exact_match_next_move_scorer; model=gemini-3.1-pro-preview_low; run=4a3M8rRWUoPphCSr2dguXH; mean accuracy; https://epoch.ai/data/benchmarks.csv	32% Reported settings Epoch AI Inspect; task 1.0.4; scorer exact_match_next_move_scorer; model=gemini-3.1-pro-preview_medium; run=WueZbBM3NwMUZPveuagrZo; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported	Not reported
Epoch · OTIS Mock AIME 2024-2025 · v1.0.5Epoch AI · Epoch AI Inspect; task 1.0.5; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	Not reported	Not reported	Not reported	95.6% Reported settings Epoch AI Inspect; task 1.0.5; scorer model_extracted_exact_match; model=gemini-3.1-pro-preview; run=bXgZZgzRE3wXoSRUgrbsVf; mean accuracy; https://epoch.ai/data/benchmarks.csv Run-specific qualification requires review: In the initial run, no epochs completed for 2/45 samples. Despite several re-runs of these two samples, no epochs completed. All epochs for the other 43/45 samples were correct. The model's score was edited to be 43/45. Source label needs review
Epoch · OTIS Mock AIME 2024-2025 · v1.1.6Epoch AI · Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	95.56% Reported settings Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=gemini-3.1-pro-preview_high; run=nhim2QCRgpir5cZdqrmUz3; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported
Epoch · SimpleQA Verified · v1.0.0Epoch AI · Epoch AI Inspect; task 1.0.0; scorer simpleqa_scorer Benchmark details Epoch AI internal evaluation. Mean accuracy under the published setting; exact task version and scorer stay separate. Related tasks share a family budget. Data by Epoch AI, CC BY. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	73.5% Reported settings Epoch AI Inspect; task 1.0.0; scorer simpleqa_scorer; model=gemini-3.1-pro-preview_high; run=VGeF8Xac4WcSAGgSXVdEBi; mean accuracy; https://epoch.ai/data/benchmarks.csv	Not reported	Not reported	Not reported
Vals AI · Finance Agent v2Vals AI · Vals FAB v2 shared six-tool harness; two-hour task limit Benchmark details Overall weighted rubric accuracy, averaged over three runs on 450 tasks. Dealbreaker failures and timeouts score zero. Related categories and all-pass scores are not additional votes. Three-model judge panel; private test set. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	42.98% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · Code Migration v1Vals AI · Vals Code Migration shared agent harness Benchmark details Hidden behavior-test pass rate, equally weighted across 40 repositories (30 CLI and 10 COBOL-to-Java). Anti-cheat failures score zero. Code-quality ratings and language splits are not additional votes. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	17.31% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · Legal Research Bench v1Vals AI · Vals Legal Research shared five-tool harness; three-hour task limit Benchmark details All-pass accuracy on 208 private legal research tasks. Every expert rubric check must pass; GPT-5.4 judges responses. Partial-credit scores and practice-area splits are not additional votes. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	20.67% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · Excel Modeling Benchmark v1Vals AI · Vals EMB shared workbook agent harness; 3.5-hour task limit Benchmark details Mean of seven equally weighted task categories. Template submissions use recalculated Excel cell accuracy; scratch submissions use rubric checks judged by GPT-5.4 Mini XHigh. Partial credit does not mean task completion. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	52.62% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · MMLU-Pro v1Vals AI · Vals MMLU-Pro five-shot chain-of-thought; regex answer grading Benchmark details Mean accuracy across 14 subjects under Vals five-shot prompting. Subject splits share one family and are not additional votes. Vals labels this benchmark archived; it is historical corroboration, not ongoing frontier evaluation. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	90.99% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-01","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65535,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · Terminal-Bench 2.1Vals AI · Vals Terminus 2 harness; pass@1 on 89 tasks Benchmark details Overall task pass rate under the shared Terminus 2 configuration. All task tests must pass. Difficulty splits share one family. Fable 5 and Opus 5 mixed-checkpoint fallback runs are excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	70.79% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported
Vals AI · Vibe Code Bench v1.1Vals AI · Vals modified OpenHands harness; five hours or 1,000 turns per app Benchmark details Mean app test pass rate on the canonical v1.1 test set. Browser Use executes workflow tests; each test requires at least 90% of its substeps. Automated judging can introduce error. Alternate agents are excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	32.03% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":"OpenHands"}	Not reported	Not reported	Not reported
Vals AI · Harvey Legal Agent Benchmark v1Vals AI · Vals Harvey held-out legal file-work protocol; internet disabled Benchmark details Mean task resolution across two judges (GPT-5.5 Medium and Sonnet 4.6), requiring every criterion to pass. Criteria pass rates and practice-area splits do not add votes. Shared Harvey family; Fable 5 fallback run excluded. Unit: percent. Source collected 2026-09-12.	Not reported	Not reported	0% Reported settings {"checkpoint":"gemini-3.1-pro-preview","evaluatorModel":"google/gemini-3.1-pro-preview","sourceUpdatedOn":"2026-09-10","reasoning_effort":"high","compute_effort":null,"reasoning":null,"temperature":1,"top_p":null,"max_output_tokens":65536,"verbosity":null,"provider":"Google","harness":null}	Not reported	Not reported	Not reported

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 · 7,424 results across 240 models · Collected 2026-09-12

Gemini 3.1 Pro effort benchmarks

Gemini 3.1 Pro

Capability score evidence

One configuration, seven capabilities

Agentic

Hard reasoning

Coding

Human pref

Knowledge

Multimodal

Long context