UnifyBench

RankingEffort benchmarks

Qwen3.7-flash

Compare models

8 measured results · 2 effort settings · 4 benchmark/harness combinations

Model evidence and sources →

Change model

Compare with

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

Compare with

Capability profile

One configuration, seven capabilities

Comparison score · 0–100
Qwen3.7-flash · Non-reasoning capability radarComparison scores from 0 to 100. Unknown capabilities have no point; lines stop at gaps. Hollow points are preliminary. Exact values and support labels follow the chart.AgenticHard reasoningCodingHuman prefKnowledgeMultimodalLong context50100Qwen3.7-flash · Non-reasoning · Hard reasoning: 11.5 · Supported

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

  • Agentic

    Unknown

  • Hard reasoning

    A11.5Supported

  • Coding

    Unknown

  • Human pref

    Unknown

  • Knowledge

    Unknown

  • 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.

Capability score evidence

Qwen3.7-flash · Non-reasoning: 4.4 overall comparison score. Meets the overall evidence requirements.

Hard reasoning only. The available task mix affects this estimate; missing areas remain unknown.

Which sources carry the weight?
  • Epoch AI · 100.0% direct comparison weight · independent of the model provider
Which benchmark families carry the weight?
  • GPQA · 30.1% direct comparison weight
  • Aime · 29.6% direct comparison weight
  • epoch-game-puzzles · 24.8% direct comparison weight
  • FrontierMath · 15.4% 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.

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 2026-10-23.

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 sourceRank before → afterRank movementCohort remaining / lost supportScore after (change)
Artificial Analysis172 → 168Up 4203 remaining / 241 lost4.0 (-0.4 points)
ARC Prize254 → 259Down 5444 remaining / 0 lost5.5 (+1.1 points)
Datacurve254 → 254No movement444 remaining / 0 lost3.5 (-0.9 points)
LiveCodeBench254 → 254No movement444 remaining / 0 lost4.4 (-0.0 points)
Scale254 → 255Down 1444 remaining / 0 lost5.1 (+0.7 points)
Terminal-Bench254 → 254No movement444 remaining / 0 lost4.4 (-0.0 points)
Cognition254 → 252Up 2444 remaining / 0 lost4.5 (+0.1 points)
LiveBench253 → 254Down 1443 remaining / 1 lost4.5 (+0.1 points)
SWE-rebench254 → 255Down 1444 remaining / 0 lost4.5 (+0.1 points)
Epoch AIComparison support lost; no rank431 remaining / 13 lostUnavailable
Vals AI254 → 254No movement443 remaining / 1 lost4.2 (-0.2 points)
Provider reports · Anthropic254 → 254No movement444 remaining / 0 lost4.4 (-0.0 points)
Provider reports · Moonshot AI254 → 254No movement444 remaining / 0 lost4.3 (-0.1 points)
Provider reports · Meta254 → 254No movement444 remaining / 0 lost4.4 (-0.0 points)
Provider reports · Mistral254 → 254No movement444 remaining / 0 lost4.4 (+0.0 points)
Provider reports · OpenAI254 → 254No movement444 remaining / 0 lost4.4 (+0.0 points)
All provider reports254 → 254No movement444 remaining / 0 lost4.3 (-0.1 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 evidenceRecomputed scoreChange
aime3.4-1.0
frontiermath5.4+1.0
arc-agi5.3+0.9
deepswe3.5-0.9
epoch-game-puzzles3.5-0.8
hle4.7+0.4
mmmu-pro4.7+0.3
gpqa4.1-0.3
simpleqa4.6+0.2
tau-bench4.2-0.2
mmlu-pro4.6+0.2
swe-rebench4.5+0.1
livebench-instructions4.5+0.1
frontiercode4.5+0.1
multi-swe-bench4.3-0.1
omniscience4.3-0.1
livebench-language4.5+0.1
livebench-coding4.3-0.1
ifbench4.5+0.1
swe-atlas-qna4.3-0.1
harvey4.3-0.1
arc-agi-34.3-0.1
gdpval4.3-0.1
apex-agents4.4+0.1
enigma-eval4.3-0.1
critpt4.3-0.0
gdp-pdf4.4+0.0
terminal-bench-science4.4+0.0
livebench-reasoning4.4-0.0
terminal-bench4.4-0.0
vibe-code4.4-0.0
analyst-agent4.4+0.0
gmmlu4.4+0.0
itbench4.4-0.0
vals-finance-agent4.4-0.0
vals-legal-research4.4-0.0
swe-atlas-test-writing4.4-0.0
briefcase4.4-0.0
automationbench4.4-0.0
aa-lcr4.4+0.0
vals-code-migration4.4-0.0
vals-tax-agent4.4-0.0
livebench-math4.4+0.0
math5004.4+0.0
vals-excel-modeling4.4-0.0
livecodebench4.4+0.0
enterprise-ops4.4+0.0
livebench-data4.4+0.0
swe-atlas-refactoring4.4-0.0
swe-bench-pro4.4+0.0
What evidence is missing from the fit?

Across all collected settings for this model: 8 observations, 4 contributing, 0 matched without graph weight and 4 excluded. Counts do not establish rank eligibility.

  • 4 · Effort is not identified
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 / sourceNon-reasoningUnspecified
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 2026-10-18.

19.3%
Reported settings

Epoch AI Inspect; task 2.0.0; scorer verification_code; model=qwen3.7-flash_none; run=doNrgGQY5SytQV22WRqw5L; mean accuracy; https://epoch.ai/data/benchmarks.csv

19.3%
Reported settings

Epoch AI Inspect; task 2.0.0; scorer verification_code; model=qwen3.7-flash; run=6pTYmQCw3EMbErvAqpivH5; 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 2026-10-18.

80.81%
Reported settings

Epoch AI Inspect; task 1.0.11; scorer choice; model=qwen3.7-flash_none; run=Yhds8GYkCXWgRQhuLiURua; mean accuracy; https://epoch.ai/data/benchmarks.csv

82.32%
Reported settings

Epoch AI Inspect; task 1.0.11; scorer choice; model=qwen3.7-flash; run=i6i3gQf3ng2aAipvowPpws; mean accuracy; https://epoch.ai/data/benchmarks.csv

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 2026-10-18.

14%
Reported settings

Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=qwen3.7-flash_none; run=i7MVHsH5pikQKjYn6LBqHS; mean accuracy; https://epoch.ai/data/benchmarks.csv

15%
Reported settings

Epoch AI Inspect; task 1.0.4; scorer agent_submitted_next_move_scorer; model=qwen3.7-flash; run=5e4Jp9dxtNisWx2wbRRZxM; mean accuracy; https://epoch.ai/data/benchmarks.csv

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 2026-10-18.

77.78%
Reported settings

Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=qwen3.7-flash_none; run=SZDgstFQJAGeGQwMXgNx6w; mean accuracy; https://epoch.ai/data/benchmarks.csv

86.67%
Reported settings

Epoch AI Inspect; task 1.1.6; scorer model_extracted_exact_match; model=qwen3.7-flash; run=TjoS7sNuqVTc4LVyGRo4wa; mean accuracy; https://epoch.ai/data/benchmarks.csv

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 · 8,916 results across 259 models · Collected 2026-10-18