Qwen3.8 27B (Xhigh) · Intelligence Index v5.0 ↗
1 more reported result
33.68 index · Qwen3.8 27B (Xhigh) · Intelligence Index v4.3.2 · Artificial Analysis ↗
Qwen / MODEL EVIDENCE
Qwen3.8 27B
54 benchmarks · 77 observations · 57 sources · released 2026-08-14
Scores at a glance
54 benchmarks · large dot: Qwen3.8 27B; small dots: closest models where reported
AA-Briefcase v1.1 ↗1392.67 Elo
Qwen3.8 27B (Xhigh) · AA-Briefcase v1.1 · overall Elo ↗
AutomationBench-AA ↗48.24%
Qwen3.8 27B (Xhigh) · AutomationBench-AA · guardrail-gated objective score ↗
GDP.pdf ↗16.40%
Qwen3.8 27B (Xhigh) · GDP.pdf · all-pass ↗
Terminal-Bench 4.0 ↗5.56%
Qwen3.8 27B (Xhigh) · Terminal-Bench 4.0 · AA harness ↗
Qwen3.8 27B (Xhigh) · Terminal-Bench-Science 0.1 · AA harness ↗
1 more reported result
0 % · Qwen 3.8 27B (Mini-SWE-agent) · Terminal-Bench-Science · Score · Vals AI · Terminal-Bench-Science ↗
FrontierCode 2.0 ↗14.91%
Qwen3.8 27B (Xhigh) · FrontierCode 2.0 · AA implementation ↗
Qwen3.8 27B (Xhigh) · CritPt v2 · hard subset · AA implementation ↗
HLE-Diamond ↗15.30%
Qwen3.8 27B (Xhigh) · HLE-Diamond · AA implementation ↗
AA-Omniscience v1.1 ↗-6.37 index
Qwen3.8 27B (Xhigh) · AA-Omniscience v1.1 · error-penalized index ↗
1 more reported result
-9.98 index · Qwen3.8 27B (Xhigh) · AA-Omniscience · error-penalized score · Artificial Analysis ↗
Qwen3.8 27B (Xhigh) · AA-Omniscience v1.1 · accuracy ↗
1 more reported result
15.58% · Qwen3.8 27B (Xhigh) · AA-Omniscience · accuracy · Artificial Analysis ↗
Qwen3.8 27B (Xhigh) · AA-Omniscience v1.1 · hallucination rate, lower is better ↗
1 more reported result
30.29% · Qwen3.8 27B (Xhigh) · AA-Omniscience · hallucination rate, lower is better · Artificial Analysis ↗
GDPval-AA v2.1 ↗1418.98 Elo
Qwen3.8 27B (Xhigh) · GDPval-AA v2.1 · Elo ↗
EnterpriseOps-Gym-AA ↗44.23%
Qwen3.8 27B (Xhigh) · EnterpriseOps-Gym-AA ↗
FrontierCode 1.1 ↗34.23%
Qwen3.8 27B (Xhigh) · FrontierCode 1.1 · AA implementation ↗
AA-LCR v1.1 ↗82.00%
Qwen3.8 27B (Xhigh) · AA-LCR v1.1 ↗
SciCode ↗46.64%
Qwen3.8 27B (Xhigh) · SciCode · AA implementation ↗
4 more reported results
46.64% · Qwen3.8 27B (xhigh) · SciCode · Score · % · Epoch AI · SciCode ↗
40.05% · Qwen3.8 27B (low) · SciCode · Score · % · Epoch AI · SciCode ↗
39% · Qwen3.8 27B (medium) · SciCode · Score · % · Epoch AI · SciCode ↗
36.23% · Qwen3.8 27B (no thinking) · SciCode · Score · % · Epoch AI · SciCode ↗
Humanity’s Last Exam ↗33.92%
Qwen3.8 27B (Xhigh) · Humanity’s Last Exam · AA implementation ↗
GPQA Diamond ↗88.89 %
Qwen 3.8 27B · GPQA Diamond · Accuracy ↗
1 more reported result
90.51% · Qwen3.8 27B (Xhigh) · GPQA Diamond · AA implementation · Artificial Analysis ↗
MMMU-Pro ↗83.93 %
Qwen 3.8 27B · MMMU Pro · Accuracy ↗
1 more reported result
76.30% · Qwen3.8 27B (Xhigh) · MMMU-Pro · AA implementation · Artificial Analysis ↗
MLCR-AA ↗22.78%
Qwen3.8 27B (Xhigh) · MLCR-AA · overall ↗
CritPt v2 ↗24.00%
Qwen3.8 27B (Medium) · CritPt v2 · code tools ↗
τ³-Banking ↗48.04%
Qwen3.8 27B (Xhigh) · τ³-Banking ↗
Terminal-Bench 2.1 ↗79.78%
Qwen3.8 27B (Xhigh) · Terminal-Bench 2.1 ↗
CritPt ↗5.43%
Qwen3.8 27B (Xhigh) · CritPt · original ↗
4 more reported results
5.43% · Qwen3.8 27B (xhigh) · CritPt · Accuracy · % · Epoch AI · CritPt ↗
0.29% · Qwen3.8 27B (no thinking) · CritPt · Accuracy · % · Epoch AI · CritPt ↗
0% · Qwen3.8 27B (medium) · CritPt · Accuracy · % · Epoch AI · CritPt ↗
0% · Qwen3.8 27B (low) · CritPt · Accuracy · % · Epoch AI · CritPt ↗
ErdősBench ↗2.800 / 4
Qwen3.8-27B · ErdősBench · mean progress / 4 · API ↗
ARC-AGI-1 ↗87.5%
Qwen3.8 27B (xhigh) · ARC-AGI · Score · % ↗
3 more reported results
69.17% · Qwen3.8 27B (low) · ARC-AGI · Score · % · Epoch AI · ARC-AGI ↗
68.67% · Qwen3.8 27B (medium) · ARC-AGI · Score · % · Epoch AI · ARC-AGI ↗
34% · Qwen3.8 27B (no thinking) · ARC-AGI · Score · % · Epoch AI · ARC-AGI ↗
Qwen3.8 27B · APEX-Agents · Pass@1 score · % ↗
ARC-AGI-2 ↗42.36%
Qwen3.8 27B (xhigh) · ARC-AGI-2 · Score · % ↗
3 more reported results
22.78% · Qwen3.8 27B (low) · ARC-AGI-2 · Score · % · Epoch AI · ARC-AGI-2 ↗
13.19% · Qwen3.8 27B (medium) · ARC-AGI-2 · Score · % · Epoch AI · ARC-AGI-2 ↗
1.53% · Qwen3.8 27B (no thinking) · ARC-AGI-2 · Score · % · Epoch AI · ARC-AGI-2 ↗
Qwen3.8 27B (xhigh) · Surface Evolver Bench · Mean score · % ↗
ProofBench ↗16 %
Qwen 3.8 27B · ProofBench · Accuracy ↗
1 more reported result
16% · Qwen3.8 27B · ProofBench · Accuracy · % · Epoch AI · ProofBench ↗
LMCA ↗41.36%
Qwen3.8 27B · LMCA · Score · % ↗
DTBench ↗88%
Qwen3.8 27B · DTBench · Accuracy · % ↗
Text Arena (Coding) ↗1593.29 Elo
Qwen3.8 27B · WebDev Arena · Elo ↗
Epoch Capabilities Index (ECI) ↗151.69 ECI
Qwen 3.8 27B · General ECI · 26 October 2026 fit ↗
Legal Research Bench ↗36.06 %
Qwen 3.8 27B · Legal Research Bench · All-Pass ↗
LegalBench ↗82.43 %
Qwen 3.8 27B · LegalBench · Accuracy ↗
Qwen 3.8 27B · Harvey's Legal Agent Benchmark v1 · Accuracy ↗
Tax Agent Bench ↗30.89 %
Qwen 3.8 27B · Tax Agent Bench · Accuracy ↗
Vals Finance Agent v2 ↗48.55 %
Qwen 3.8 27B · Finance Agent v2 · Accuracy ↗
Qwen 3.8 27B · Excel Modeling Benchmark v1 · Accuracy ↗
MortgageTax & TaxEval v2 ↗64.94 %
Qwen 3.8 27B · MortgageTax · Accuracy ↗
1 more reported result
70.85 % · Qwen 3.8 27B · TaxEval · Accuracy · Vals AI · TaxEval ↗
MedCode ↗28.7 %
Qwen 3.8 27B · MedCode · Accuracy ↗
MedScribe ↗83.85 %
Qwen 3.8 27B · MedScribe · Accuracy ↗
MMLU-Pro ↗84.34 %
Qwen 3.8 27B · MMLU Pro · Accuracy ↗
Qwen 3.8 27B · SAGE · Accuracy ↗
Vibe Code Bench ↗64.85 %
Qwen 3.8 27B (OpenHands) · Vibe Code Bench · original task set ↗
Qwen 3.8 27B · ProgramBench · Accuracy ↗
SkillsBench ↗38.11 %
Qwen 3.8 27B (OpenHands) · SkillsBench · Score ↗
IOI v1 & IOI v2 ↗39.06 %
Qwen 3.8 27B · IOI · Accuracy ↗
Code Migration v1.1 ↗14.16 %
Qwen 3.8 27B · Code Migration v1 · Accuracy ↗
Qwen 3.8 27B (Mini-SWE-agent) · SWE-bench Verified · Score ↗
LiveCodeBench ↗84 %
Qwen 3.8 27B · LiveCodeBench · Accuracy ↗
CyberBench v1.1 ↗67.8 %
Qwen 3.8 27B · CyberBench · Accuracy ↗
Closest models by score
Similar positions on shared benchmark leaderboards. Smaller gaps are closer.
41 of 44 shared benchmark comparisons
Step 5 Preview →StepFun
28 of 44 shared benchmark comparisons
GPT-6 Luna →OpenAI
32 of 44 shared benchmark comparisons
How similarity is measured
Mean absolute gap in percentile positions on shared benchmark settings; lower is closer. At least 60% overlap is required, with a minimum of three when available. Independent evaluations are used when there are at least three; otherwise published report comparisons are used. Repeated model observations within one setting are averaged before ranking. Writing-style traits and composite indices are excluded. This describes similarity, not an overall quality ranking.
Sources
Every plotted observation links to its evaluation. Reported configurations, benchmark versions and task subsets can differ; similarity describes shared scores, not training lift.
- ARC Prize - Leaderboard ↗
- Arena leaderboard — WebDev evaluation ↗
- AA-Briefcase v1.1: Agentic Knowledge Work Benchmark | Artificial Analysis ↗
- Artificial Analysis Intelligence Index v5.0 — Artificial Analysis ↗
- AA-LCR v1.1: Long Context Reasoning Benchmark | Artificial Analysis ↗
- AutomationBench-AA: Agentic SaaS Workflow Benchmark | Artificial Analysis ↗
- CritPt Benchmark | Artificial Analysis ↗
- CritPt v2 Benchmark | Artificial Analysis ↗
- CritPt v2 — Hard Subset — Artificial Analysis ↗
- EnterpriseOps-Gym-AA Benchmark | Artificial Analysis ↗
- FrontierCode 1.1 Benchmark | Artificial Analysis ↗
- FrontierCode 2.0 Benchmark — Artificial Analysis ↗
- GDP.pdf Benchmark | Artificial Analysis ↗
- GDPval-AA v2.1 Benchmark | Artificial Analysis ↗
- GPQA Diamond Benchmark | Artificial Analysis ↗
- HLE-Diamond Benchmark — Artificial Analysis ↗
- Humanity's Last Exam Benchmark | Artificial Analysis ↗
- MLCR-AA: Medical Long Context Reasoning Benchmark | Artificial Analysis ↗
- MMMU-Pro Benchmark | Artificial Analysis ↗
- AA-Omniscience v1.1: Knowledge and Hallucination Benchmark — Artificial Analysis ↗
- SciCode Benchmark | Artificial Analysis ↗
- 𝜏³-Banking Benchmark | Artificial Analysis ↗
- Terminal-Bench-Science 0.1 Benchmark | Artificial Analysis ↗
- Terminal-Bench 2.1 Benchmark | Artificial Analysis ↗
- Terminal-Bench 4.0 Benchmark | Artificial Analysis ↗
- DTBench — Conceptual Reasoning Index ↗
- LMCA — Conceptual Reasoning Index ↗
- APEX-Agents ↗
- ARC-AGI-2 ↗
- CritPt v1 ↗
- AI benchmarks & capabilities ↗
- ErdősBench: research mathematics, scoped progress and proof audits ↗
- Code Migration v1 — evaluation methodology and results ↗
- CyberBench — evaluation methodology and results ↗
- Excel Modeling Benchmark v1 — evaluation methodology and results ↗
- Finance Agent v2 — evaluation methodology and results ↗
- GPQA Diamond — evaluation methodology and results ↗
- Harvey's Legal Agent Benchmark v1 — evaluation methodology and results ↗
- IOI — evaluation methodology and results ↗
- LiveCodeBench — evaluation methodology and results ↗
- LegalBench — evaluation methodology and results ↗
- Legal Research Bench — evaluation methodology and results ↗
- MedCode — evaluation methodology and results ↗
- MedScribe — evaluation methodology and results ↗
- MMLU Pro — evaluation methodology and results ↗
- MMMU Pro — evaluation methodology and results ↗
- MortgageTax — evaluation methodology and results ↗
- ProgramBench — evaluation methodology and results ↗
- ProofBench — evaluation methodology and results ↗
- SAGE — evaluation methodology and results ↗
- SkillsBench — evaluation methodology and results ↗
- SWE-bench Verified — evaluation methodology and results ↗
- Tax Agent Bench — evaluation methodology and results ↗
- TaxEval — evaluation methodology and results ↗
- Terminal-Bench-Science — evaluation methodology and results ↗
- Vibe Code Bench — evaluation methodology and results ↗
- Surface Evolver Bench ↗
Names used in source tables
Qwen 3.8 27B · Qwen 3.8 27B (Mini-SWE-agent) · Qwen 3.8 27B (OpenHands) · Qwen3.8 27B · Qwen3.8 27B (Medium) · Qwen3.8 27B (Xhigh) · Qwen3.8 27B (low) · Qwen3.8 27B (medium) · Qwen3.8 27B (no thinking) · Qwen3.8 27B (xhigh) · Qwen3.8-27B