Model comparison
GPT-5.4 vs Qwen3.5-27B
Head-to-head evidence from 17 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: GPT-5.4 #8 (Supported); Qwen3.5-27B #45 (Supported). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. GPT-5.4 and Qwen3.5-27B share 17 comparable benchmark results. 3 of 8 categories are comparable. 35 results are unique to GPT-5.4; 11 to Qwen3.5-27B.
Updated July 23, 2026- Shared results
- 17
- GPT-5.4 only
- 35
- Qwen3.5-27B only
- 11
- Comparable categories
- 3 / 8
Pick GPT-5.4 if you want the stronger benchmark profile. Qwen3.5-27B only becomes the better choice if knowledge is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 17 shared benchmark results across 6 evidence categories; 3 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
GPT-5.4 is clearly ahead on the BenchAlign aggregate, 74.24 to 60.7. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
GPT-5.4's sharpest advantage is in agentic, where it averages 77.2 against 52. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 75.1% to 41.6%. Qwen3.5-27B does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.
GPT-5.4 is also the more expensive model on tokens at $2.50 input / $15.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Qwen3.5-27B. That is roughly Infinityx on output cost alone. GPT-5.4 gives you the larger context window at 1.05M, compared with 262K for Qwen3.5-27B.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | GPT-5.4 | Δ | Qwen3.5-27B |
|---|---|---|---|
| Agentic | GPT-5.477.2 | Margin← 25.2 | Qwen3.5-27B52.0 |
| Knowledge | GPT-5.457.6 | Margin→ 25.1 | Qwen3.5-27B82.7 |
| Coding | GPT-5.457.7 | Margin→ 7.2 | Qwen3.5-27B64.9 |
| Reasoning | GPT-5.4Not measured | MarginNo overlap | Qwen3.5-27B60.6 |
| Math | GPT-5.442.5 | MarginNo overlap | Qwen3.5-27BNot measured |
| Multilingual | GPT-5.4Not measured | MarginNo overlap | Qwen3.5-27B82.2 |
| Multimodal | GPT-5.473.2 | MarginNo overlap | Qwen3.5-27BNot measured |
| Inst. Following | GPT-5.4Not measured | MarginNo overlap | Qwen3.5-27B95.0 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
Terminal-Bench 2.0
AgenticA 75.1%B 41.6%Winner: GPT-5.4Δ 33.5Terminal-Bench 2.0: GPT-5.4 scored 75.1%; Qwen3.5-27B scored 41.6%. GPT-5.4 wins this benchmark. - Source ↗
BrowseComp
AgenticA 82.7%B 61%Winner: GPT-5.4Δ 21.7BrowseComp: GPT-5.4 scored 82.7%; Qwen3.5-27B scored 61%. GPT-5.4 wins this benchmark. - Source ↗
OSWorld-Verified
AgenticA 75%B 56.2%Winner: GPT-5.4Δ 18.8OSWorld-Verified: GPT-5.4 scored 75%; Qwen3.5-27B scored 56.2%. GPT-5.4 wins this benchmark. - Source ↗
GPQA
KnowledgeA 92.8%B 85.5%Winner: GPT-5.4Δ 7.3GPQA: GPT-5.4 scored 92.8%; Qwen3.5-27B scored 85.5%. GPT-5.4 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | GPT-5.4 | Qwen3.5-27B | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | GPT-5.4$2.5 input / $15 output | Qwen3.5-27B$0 input / $0 output | Qwen3.5-27B has the lower combined listed price. |
| Generation speedtokens per second | GPT-5.474 tok/s | Qwen3.5-27BNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | GPT-5.4151.79 s | Qwen3.5-27BNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | GPT-5.41.05M | Qwen3.5-27B262K | GPT-5.4 lists the larger context window. |
Benchmark Deep Dive
AgenticGPT-5.4 wins17 benchmarks
| Benchmark | GPT-5.4 | Qwen3.5-27B | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 75.1% | 41.6% | GPT-5.4 leads |
| CyberGymSource | 79.0% | — | Not comparable |
| BrowseCompSource | 82.7% | 61% | GPT-5.4 leads |
| OSWorld-VerifiedSource | 75% | 56.2% | GPT-5.4 leads |
| MCP AtlasSource | 70.6% | — | Not comparable |
| ToolathlonSource | 54.6% | — | Not comparable |
| τ²-bench resultsSource | 87.1% | 93.9% | Qwen3.5-27B leads |
| Claw-EvalSource | 60.3% | — | Not comparable |
| DeepSearchQASource | 73.6% | — | Not comparable |
| AA Agentic IndexSource | 41.1% | — | Not comparable |
| APEX-Agents-AASource | 33.3% | — | Not comparable |
| GDPval-AASource | 44.7% | — | Not comparable |
| GDPval-AASource | 1395 | — | Not comparable |
| Gert LabsSource | 64.89% | 39.41% | GPT-5.4 leads |
| ResearchClawBenchSource | 15.3% | — | Not comparable |
| JobBenchSource | 38.9% | — | Not comparable |
| ExploitGymSource | 6.0% | — | Not comparable |
CodingQwen3.5-27B wins8 benchmarks
| Benchmark | GPT-5.4 | Qwen3.5-27B | Result |
|---|---|---|---|
| LiveCodeBench ProSource | 87.5% | — | Not comparable |
| SWE-bench ProSource | 57.7% | — | Not comparable |
| React Native EvalsSource | 85.3% | — | Not comparable |
| Vibe Code BenchSource | 67.42% | — | Not comparable |
| AA Coding IndexSource | 71.0% | — | Not comparable |
| AA-SciCodeSource | 56.6% | 39.5% | GPT-5.4 leads |
| SWE-bench VerifiedSource | — | 72.4% | Not comparable |
| SWE-RebenchSource | — | 58.9% | Not comparable |
Reasoning3 benchmarks
KnowledgeQwen3.5-27B wins15 benchmarks
| Benchmark | GPT-5.4 | Qwen3.5-27B | Result |
|---|---|---|---|
| GPQASource | 92.8% | 85.5% | GPT-5.4 leads |
| HLESource | 52.1% | — | Not comparable |
| HLE w/o toolsSource | 39.8% | — | Not comparable |
| GPQA-DSource | 92.8% | — | Not comparable |
| HealthBench HardSource | 40.1% | — | Not comparable |
| MedXpertQA (Text)Source | 59.6% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 51.4% | 33.8% | GPT-5.4 leads |
| AA-GPQA DiamondSource | 92.0% | 85.8% | GPT-5.4 leads |
| AA-HLESource | 41.6% | 22.2% | GPT-5.4 leads |
| AA-Omniscience IndexSource | 5.7% | -42.0% | GPT-5.4 leads |
| AA-Omniscience AccuracySource | 50.0% | 21.0% | GPT-5.4 leads |
| AA-Omniscience Hallucination RateSource | 88.6% | 79.7% | Qwen3.5-27B leads |
| HealthBench ProfessionalSource | 48.1% | — | Not comparable |
| MMLU-ProSource | — | 86.1% | Not comparable |
| SuperGPQASource | — | 65.6% | Not comparable |
Math2 benchmarks
Multilingual1 benchmarks
| Benchmark | GPT-5.4 | Qwen3.5-27B | Result |
|---|---|---|---|
| MMLU-ProXSource | — | 82.2% | Not comparable |
Multimodal15 benchmarks
| Benchmark | GPT-5.4 | Qwen3.5-27B | Result |
|---|---|---|---|
| MMMU-ProSource | 81.2% | — | Not comparable |
| OfficeQA ProSource | 53.2% | — | Not comparable |
| MMMU-Pro w/ PythonSource | 82.1% | — | Not comparable |
| CharXivSource | 82.8% | — | Not comparable |
| ERQASource | 65.4% | — | Not comparable |
| SimpleVQASource | 61.1% | — | Not comparable |
| ScreenSpot ProSource | 85.4% | — | Not comparable |
| ZeroBenchSource | 41.0% | — | Not comparable |
| MedXpertQA (MM)Source | 77.1% | — | Not comparable |
| AA-MMMU-ProSource | 78.4% | 75.0% | GPT-5.4 leads |
| Design Arena WebsiteSource | 1250 | — | Not comparable |
| MMMUSource | — | 82.3% | Not comparable |
| MMVUSource | — | 73.3% | Not comparable |
| MathVisionSource | — | 86.0% | Not comparable |
| V*Source | — | 93.7% | Not comparable |
Frequently Asked Questions (4)
Which is better, GPT-5.4 or Qwen3.5-27B?
GPT-5.4 is ahead on BenchLM's BenchAlign leaderboard, 74.24 to 60.7. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 75.1% and 41.6%.
Which is better for knowledge tasks, GPT-5.4 or Qwen3.5-27B?
Qwen3.5-27B has the edge for knowledge tasks in this comparison, averaging 82.7 versus 57.6. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.
Which is better for coding, GPT-5.4 or Qwen3.5-27B?
Qwen3.5-27B has the edge for coding in this comparison, averaging 64.9 versus 57.7. Inside this category, AA-SciCode is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, GPT-5.4 or Qwen3.5-27B?
GPT-5.4 has the edge for agentic tasks in this comparison, averaging 77.2 versus 52. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.