Across 33 comparable benchmarks, DeepSeek-V4.1-Flash-1017 has better reported results on 30 and worse on 3. 1 further shared result is listed below, each with the reason it is not counted.
Moderate evidence33 of 34 shared results countedHow we compare
Price basis: Listed rates vs peak-hour rates. The price section below shows the applicable periods.
| Fact | Qwen3.8-27B | DeepSeek-V4.1-Flash-1017 |
|---|---|---|
| Context window | 256K tokens | 1M tokens |
| Input / 1M | $0.500 | $0.300 |
| Output / 1M | $3.00 | $1.20 |
| Blended (4:1) | $0.712 | $0.268 |
| Status | Available | Available |
| Released | August 14, 2026 | October 17, 2026 |
Blended is 4:1 input to output with 90% cache hits at listed rates — set your own mix.
Mean of 18 shared benchmarks grouped by capability on a 0–100 scale. How averages are built
DeepSeek-V4.1-Flash-1017 ahead on 8 of 8 counted
DeepSeek-V4.1-Flash-1017 ahead on 1 of 1 counted
DeepSeek-V4.1-Flash-1017 ahead on 3 of 3 counted
DeepSeek-V4.1-Flash-1017 ahead on 3 of 3 counted
DeepSeek-V4.1-Flash-1017 ahead on 5 of 7 counted
DeepSeek-V4.1-Flash-1017 ahead on 6 of 7 counted
No shared comparable results for Mathematics, Research & search, Long context, Cybersecurity.
Published scores, with matching modes where available. Gaps show Qwen3.8-27B relative to DeepSeek-V4.1-Flash-1017. Reading the results
| Benchmark | Qwen3.8-27B | DeepSeek-V4.1-Flash-1017 | Gap |
|---|---|---|---|
| AA Cyber IndexRepository Engineering · Composite index (0–100) | 12.7Extra-High · tools | 56.8≠Max · tools | 0.22×#21 vs #2 of 22 |
| Code MigrationMaintenance & Optimization · Hidden-test pass rate (%) | 14.2Extra-High · tools | 44.2≠Max · tools | −30.1 pts#59 vs #26 of 79 |
| DeepSWERepository Engineering · Pass@1 (DeepSWE v1.1) | 42.2Thinking · tools | 77.0≠Unspecified Reasoning · tools | −34.8 pts#64 vs #4 of 73 |
| FrontierCode 2.0Repository Engineering · Weighted rubric score | 14.9Thinking · tools | 19.3≠Unspecified Reasoning · tools | 0.77×#27 vs #22 of 36 |
| NL2Repo-BenchRepository Engineering · Average test pass rate | 42.3Thinking · tools | 68.2≠Unspecified Reasoning · tools | −25.9 pts#18 vs #2 of 21 |
| Terminal-Bench 2.1Agentic Development · Accuracy | 73.0Thinking · tools | 91.0≠Unspecified Reasoning · tools | −18.0 pts#46 vs #4 of 68 |
| Terminal-Bench 4.0Agentic Development · Resolution rate (%) | 5.6Extra-High · tools | 42.3≠Unspecified Reasoning · tools | −36.7 pts#67 vs #25 of 90 |
| Vibe Code Bench v1.1Code Generation & Editing · Pass rate (%) | 64.8Extra-High · tools | 87.7≠Max · tools | −22.9 pts#54 vs #15 of 121 |
| ARC-AGI-1Abstract Generalization · Score (%) | 87.5Extra-High · no tools | 94.5≠Thinking High · no tools | −7.0 pts#40 vs #21 of 87 |
| ARC-AGI-2Abstract Generalization · Score (% solved); cost per task (USD)Not counted · Speed and cost telemetry is reported, not scored | 42.4Extra-High · no tools | 78.1≠Max · no tools | 0.54×#48 vs #22 of 80 |
| CritPtScientific Reasoning · Score | 5.4Extra-High · no tools | 24.0≠Max · no tools | 0.23×#68 vs #22 of 134 |
| GPQA DiamondScientific Reasoning · Accuracy | 89.2Thinking · no tools | 93.2≠Max · no tools | −4.0 pts#58 vs #34 of 200 |
| SciCodeScientific Computing · Score | 46.6Extra-High · no tools | 55.9≠Max · no tools | 0.83×#58 vs #29 of 98 |
| Agents' Last Exam v1.0Tool Orchestration · Pass rate | 20.4Thinking · tools· pass-rate | 35.4≠Unspecified Reasoning · tools· pass-rate | −15.0 pts#24 vs #4 of 34 |
| APEX-AgentsTool Orchestration · Score | 47.5Reported best | 63.4≠Max · tools | 0.75×#37 vs #15 of 61 |
| AutomationBench-AATool Orchestration · Guardrail-safe objective completion rate | 48.2Thinking · tools | 66.6≠Max · tools | −18.4 pts#15 vs #12 of 18 |
| BabyVisionVisual Understanding · Accuracy | 85.6Thinking · tools | 92.5≠Unspecified Reasoning · tools | −6.9 pts#13 vs #8 of 17 |
| ChartographyDocuments & Charts · Score | 52.1Thinking · tools | 83.0≠Unspecified Reasoning · tools | 0.63×#24 vs #9 of 29 |
| CharXiv RQDocuments & Charts · Score | 90.2Thinking · tools | 85.6≠Max · tools | 1.05×#7 vs #14 of 21 |
| GDP.pdfDocuments & Charts · Task pass rate (%) | 16.4Extra-High · no tools | 17.2≠Max · no tools | −0.8 pts#39 vs #37 of 96 |
| MMMU-ProVisual Understanding · Accuracy | 76.3Extra-High · no tools | 82.5≠Max · tools | −6.2 pts#58 vs #18 of 136 |
| PerceptionBenchVisual Understanding · Accuracy | 52.4Thinking · no tools | 47.7≠Max · no tools | +4.7 pts#14 vs #15 of 19 |
| ZeroBench MainVisual Understanding · Pass@1 / Pass@5 / Pass^5 | 12.0Thinking · no tools | 51.0≠Unspecified Reasoning · tools | −39.0 pts#19 vs #8 of 21 |
| AA-BriefcaseOffice & Business · AA-Briefcase Elo | 1391Thinking · tools | 1570≠Max · tools | 0.89×#17 vs #14 of 20 |
| EMB (Excel Modeling)Finance · Accuracy (%) | 59.7Extra-High · tools | 61.8≠Max · tools | −2.1 pts#38 vs #34 of 78 |
| Finance Agent v2Finance · Score | 48.5Extra-High · tools | 57.0≠Max · tools | 0.85×#47 vs #16 of 75 |
| GDPval-AA v2Cross-industry Work · Elo score | 1427Thinking · tools | 1746≠Max · tools | 0.82×#17 vs #8 of 23 |
| Harvey's Legal Agent BenchmarkLegal · Accuracy (%) | 11.3Extra-High · tools | 9.2≠Max · tools | +2.1 pts#11 vs #15 of 52 |
| Legal Research BenchLegal · Score | 36.1Extra-High · tools | 45.2≠Max · tools | 0.80×#49 vs #21 of 91 |
| Tax Agent BenchFinance · Accuracy (%, rubric × citation quality) | 30.9Extra-High · tools | 44.4≠Max · tools | −13.5 pts#41 vs #26 of 74 |
| HLEKnowledge Exams · Accuracy | 30.8Thinking · no tools | 58.8≠Max · tools | −28.0 pts#86 vs #17 of 151 |
| HLE-DiamondKnowledge Exams · Accuracy (%) | 15.3Thinking · no tools | 21.6≠Max · no tools | −6.3 pts#12 vs #11 of 16 |
| AA Intelligence Index v5.0.1Capability Indices · Index (0-100) | 20.1Extra-High · tools | 32.6≠Max · task tools | 0.62×#43 vs #25 of 62 |
| LiveBench 2.0Cross-capability Suites · Overall score (equal-weight mean of 7 suite scores, 0-100) | 62.7Max · task tools | 79.2Max · task tools | 0.79×#36 vs #10 of 51 |
Repository Engineering · Composite index (0–100)
Maintenance & Optimization · Hidden-test pass rate (%)
Repository Engineering · Pass@1 (DeepSWE v1.1)
Repository Engineering · Weighted rubric score
Repository Engineering · Average test pass rate
Agentic Development · Accuracy
Agentic Development · Resolution rate (%)
Code Generation & Editing · Pass rate (%)
Scientific Reasoning · Score
Scientific Reasoning · Accuracy
Scientific Computing · Score
Tool Orchestration · Pass rate
Tool Orchestration · Score
Tool Orchestration · Guardrail-safe objective completion rate
Visual Understanding · Accuracy
Documents & Charts · Score
Documents & Charts · Score
Documents & Charts · Task pass rate (%)
Visual Understanding · Accuracy
Visual Understanding · Accuracy
Visual Understanding · Pass@1 / Pass@5 / Pass^5
Office & Business · AA-Briefcase Elo
Finance · Accuracy (%)
Finance · Score
Cross-industry Work · Elo score
Legal · Accuracy (%)
Legal · Score
Finance · Accuracy (%, rubric × citation quality)
Knowledge Exams · Accuracy
Knowledge Exams · Accuracy (%)
Capability Indices · Index (0-100)
Cross-capability Suites · Overall score (equal-weight mean of 7 suite scores, 0-100)
A dash means no result is available.
Time-of-day providers use peak-hour rates; fixed-price providers use standard rates. Off-peak discounts remain in the published details.
Drag the handle above the bar for cache hits, the one below it for the input : output mix · 5% steps
Blended price = effective input rate at 90% cache hits × 80% + output rate × 20%
Qwen3.8-27B costs165%morethan DeepSeek-V4.1-Flash-1017 per million tokens — 2.7× the price.Using listed rates vs peak-hour rates at this token mix.
Qwen3.8-27B: International pricing. context_length <= 1000000 Alibaba Model Studio International (Singapore)
DeepSeek-V4.1-Flash-1017: Peak-hour pricing. DeepSeek official USD pricing (peak: 01:00–04:00 and 06:00–10:00 UTC Mon–Fri exc. CN holidays) DeepSeek official USD pricing (peak output)
| Price / 1M tokens | Qwen3.8-27B | DeepSeek-V4.1-Flash-1017 | Ratio |
|---|---|---|---|
| Input | $0.50 | $0.30 | 1.67× |
| Output | $3.00 | $1.20 | 2.50× |
| Cached input | $0.10 | $0.006 | 16.7× |
| Cache write | — | — | — |
| Blended | $0.712 | $0.268 | 2.65× |
Estimates use the selected token mix. Cache writes, tool fees and long-context surcharges are excluded.
Reported cost per task, paired with the score from the same run. Settings and sources may differ; costs do not affect the verdict.
Qwen3.8-27B relative to DeepSeek-V4.1-Flash-1017
Summary: 4 of 5 shared benchmarks have matching recorded settings and sources. The average is a geometric mean of 4 positive-cost ratios, one equal weight per benchmark. This benchmark set can include related tests and indexes; it is not a workload cost estimate.
Qwen3.8-27B 8.9× more expensive
Qwen3.8-27B 7× more expensive
Qwen3.8-27B 5.8× more expensive
Qwen3.8-27B 4× more expensive
Settings or sources differ
Qwen3.8-27B 8.9× more expensive
Qwen3.8-27B 7× more expensive
Qwen3.8-27B 5.8× more expensive
Qwen3.8-27B 4× more expensive
Settings or sources differ
Effort, protocol, aggregation and source match; other run conditions may differ.
Qwen3.8-27B
DeepSeek-V4.1-Flash-1017
Effort, protocol, aggregation and source match; other run conditions may differ.
Qwen3.8-27B
DeepSeek-V4.1-Flash-1017
Effort, protocol, aggregation and source match; other run conditions may differ.
Qwen3.8-27B
DeepSeek-V4.1-Flash-1017
Effort, protocol, aggregation and source match; other run conditions may differ.
Qwen3.8-27B
DeepSeek-V4.1-Flash-1017
Settings or sources differ or are not recorded; interpret the comparison with care.
Qwen3.8-27B
DeepSeek-V4.1-Flash-1017
| Benchmark | Qwen3.8-27B | DeepSeek-V4.1-Flash-1017 |
|---|---|---|
| AA Intelligence Index v4.3 | Not reported | |
| AA-Briefcase | Not reported | |
| AA-LCR | Not reported | |
| AA-Omniscience | Not reported | |
| Agent Arena | Not reported | $0.1402 / task Median · Net improvement index (percentage points vs. average model): 2 Max (With Tools + Internet) |
| Agent Arena Index | Not reported | $0.1423 / task Median · Agent Arena Index rating (index points): 1109 Max (With Tools + Internet) |
| AutomationBench-AA | Not reported | |
| Code Migration | Not reported | |
| CritPt | Not reported | |
| CursorBench 4.0 | Not reported | $0.77 / task Average · Task score: 45.6 Thinking High (With Tools) · cursorbench-4-0-cursor-agent $0.60 / task Average · Task score: 44.9 Thinking Low (With Tools) · cursorbench-4-0-cursor-agent 1 more settings$0.80 / task Average · Task score: 46.3 Thinking Medium (With Tools) · cursorbench-4-0-cursor-agent |
| EMB (Excel Modeling) | Not reported | |
| Finance Agent v2 | Not reported | |
| FrontierCS 2 | Not reported | |
| GDP.pdf | Not reported | |
| GDPval-AA v2 | Not reported | |
| GDPx | Not reported | $0.94 / task Average · GDPx — Score: 5.4 Highest (source-reported) (With Tools) · overall/score |
| Harvey's Legal Agent Benchmark | Not reported | |
| HLE | Not reported | |
| Legal Research Bench | Not reported | |
| SciCode | Not reported | |
| SWE-sweep | Not reported | |
| Terminal-Bench 4.0 | Not reported | $0.54 / task Average · Resolution rate (%): 31.8 Unspecified Reasoning (With Tools) · vals-mini-swe |
| Terminal-Bench 4.1 | Not reported | |
| Vals Index v2.1 | Not reported | |
| Vals Index v2.2 | Not reported | |
| Vibe Code Bench v1.1 | Not reported | |
| WeirdML v3 | Not reported | $0.879 / task Average · Official score (0-1; higher is better): 0.092 Thinking High (With Tools) |
| Attribute | Qwen3.8-27B | DeepSeek-V4.1-Flash-1017 |
|---|---|---|
| Availability | Available | Available |
| Provider | Alibaba | DeepSeek-AI |
| Model type | Reasoning model | Multimodal model |
| First release | August 14, 2026 | October 17, 2026 |
| Context window | 256K tokens | 1M tokens |
| Max output | 128K tokens | 384K tokens |
| Total parameters | 27B | 552B |
| Active parameters | 27B | 16B |
| Architecture | Dense | Mixture of Experts (MoE) |
| Documented capabilities | Reasoning, Multilingual, Coding, Tool use, Function calling, Long context, Computer use | Reasoning, Multilingual, Coding, Function calling, Tool use, Structured output, Long context |
| Input modalities | Text, Image, Video | Text, Image |
| Output modalities | Text | Text |
| Weights | Open weights | Open weights |
| Licence | Apache 2.0 | MIT License |
| Commercial use | Commercial use allowed | Commercial use allowed |
These are reported-score comparisons, not controlled head-to-head tests.
What people found comparing these models, or using either one. Comments also appear on the models' own pages.
Used Qwen3.8-27B or DeepSeek-V4.1-Flash-1017? Sign in with Google to share what worked, what did not, and which API or setup you tested.
No community comments on either model yet.
Alternatives with the most benchmark overlap for this matchup.