Model comparison
Claude Opus 5 vs GPT-5.4 mini
Head-to-head evidence from 18 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: Claude Opus 5 #1 (Estimated); GPT-5.4 mini #81 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Claude Opus 5 and GPT-5.4 mini share 18 comparable benchmark results. 3 of 8 categories are comparable. 61 results are unique to Claude Opus 5; 12 to GPT-5.4 mini.
Updated July 24, 2026- Shared results
- 18
- Claude Opus 5 only
- 61
- GPT-5.4 mini only
- 12
- Comparable categories
- 3 / 8
Pick Claude Opus 5 if you want the stronger benchmark profile. GPT-5.4 mini only becomes the better choice if multimodal & grounded is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 18 shared benchmark results across 5 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
Claude Opus 5 is clearly ahead on the BenchAlign aggregate, 85.88 to 55.82. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude Opus 5's sharpest advantage is in agentic, where it averages 90.8 against 65.7. The single biggest benchmark swing on the page is HLE, 64.7% to 41.5%. GPT-5.4 mini does hit back in multimodal & grounded, so the answer changes if that is the part of the workload you care about most.
Claude Opus 5 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $0.75 input / $4.50 output per 1M tokens for GPT-5.4 mini. That is roughly 5.6x on output cost alone.
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 | Claude Opus 5 | Δ | GPT-5.4 mini |
|---|---|---|---|
| Agentic | Claude Opus 590.8 | Margin← 25.1 | GPT-5.4 mini65.7 |
| Knowledge | Claude Opus 564.7 | Margin← 16.9 | GPT-5.4 mini47.8 |
| Multimodal | Claude Opus 566.9 | Margin→ 9.7 | GPT-5.4 mini76.6 |
| Coding | Claude Opus 589.5 | MarginNo overlap | GPT-5.4 miniNot measured |
| Reasoning | Claude Opus 590.4 | MarginNo overlap | GPT-5.4 miniNot measured |
| Math | Claude Opus 5Not measured | MarginNo overlap | GPT-5.4 mini21.7 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
HLE
KnowledgeA 64.7%B 41.5%Winner: Claude Opus 5Δ 23.2HLE: Claude Opus 5 scored 64.7%; GPT-5.4 mini scored 41.5%. Claude Opus 5 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Claude Opus 5 | GPT-5.4 mini | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Claude Opus 5$5 input / $25 output | GPT-5.4 mini$0.75 input / $4.5 output | GPT-5.4 mini has the lower combined listed price. |
| Generation speedtokens per second | Claude Opus 5Not available | GPT-5.4 mini201 tok/s | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Claude Opus 5Not available | GPT-5.4 mini3.85 s | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Claude Opus 5Not available | GPT-5.4 mini400K | A complete context comparison is not available. |
Benchmark Deep Dive
AgenticClaude Opus 5 wins29 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| FrontierBench v0.1Source | 43.3% | — | Not comparable |
| BrowseCompSource | 90.8% | — | Not comparable |
| HLE w/ toolsSource | 64.7% | — | Not comparable |
| DeepSearchQASource | 95.0% | — | Not comparable |
| DRACOSource | 88.6% | — | Not comparable |
| BrowseComp (10-agent, prerelease)Source | 93.6% | — | Not comparable |
| OSWorld 2.0Source | 70.6% | — | Not comparable |
| MCP AtlasSource | 85.8% | 57.7% | Claude Opus 5 leads |
| MCP-Atlas claim coverageSource | 89.1% | — | Not comparable |
| LAB all-pass (Anthropic harness)Source | 23.58% | — | Not comparable |
| LAB criterion-pass (Anthropic harness)Source | 93.74% | — | Not comparable |
| LAB all-pass (Harvey held-out)Source | 11.7% | — | Not comparable |
| LAB criterion-pass (Harvey held-out)Source | 94.1% | — | Not comparable |
| GDPval-AASource | 1861 | 1169 | Claude Opus 5 leads |
| Toolathlon-VerifiedSource | 80.6% | — | Not comparable |
| Toolathlon Verified Pass@3Source | 87.0% | — | Not comparable |
| Toolathlon Verified Pass³Source | 73.1% | — | Not comparable |
| Toolathlon Verified avg. turnsSource | 23.5 turns | — | Not comparable |
| AutomationBenchSource | 26.0% | — | Not comparable |
| AA Agentic IndexSource | 55.3% | 30.2% | Claude Opus 5 leads |
| GDPval-AASource | 68.0% | 33.5% | Claude Opus 5 leads |
| AA Tau3 BankingSource | 30.3% | — | Not comparable |
| AA BriefcaseSource | 1720 | — | Not comparable |
| aaTerminalBench21Source | 89.1% | — | Not comparable |
| Terminal-Bench 2.0Source | — | 60% | Not comparable |
| OSWorld-VerifiedSource | — | 72.1% | Not comparable |
| ToolathlonSource | — | 42.9% | Not comparable |
| τ²-bench resultsSource | — | 83.3% | Not comparable |
| APEX-Agents-AASource | — | 28.2% | Not comparable |
Coding13 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 96% | — | Not comparable |
| SWE-bench ProSource | 79.2% | — | Not comparable |
| SWE MultilingualSource | 89.5% | — | Not comparable |
| SWE MultimodalSource | 59.4% | — | Not comparable |
| deepSweSource | 68.8% | — | Not comparable |
| FrontierCode 1.1 MainSource | 53.4% | 27.0% | Claude Opus 5 leads |
| FrontierCode 1.1 ExtendedSource | 63.6% | — | Not comparable |
| ProgramBench (episode 1)Source | 83.0% | — | Not comparable |
| ProgramBenchSource | 93.0% | — | Not comparable |
| cursorBench32Source | 70.0% | — | Not comparable |
| AA Coding IndexSource | 78.0% | 56.1% | Claude Opus 5 leads |
| AA-SciCodeSource | 55.7% | 49.9% | Claude Opus 5 leads |
| Vibe Code BenchSource | — | 47.97% | Not comparable |
Reasoning5 benchmarks
KnowledgeClaude Opus 5 wins22 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| HLESource | 64.7% | 41.5% | Claude Opus 5 leads |
| HLE w/o toolsSource | 56.3% | 28.2% | Claude Opus 5 leads |
| HealthBench (raw)Source | 67.1% | — | Not comparable |
| HealthBench (length-adjusted)Source | 57.8% | — | Not comparable |
| HealthBench ProfessionalSource | 59.8% | — | Not comparable |
| HealthBench Professional (raw)Source | 73.4% | — | Not comparable |
| BioMysteryBench (human-solvable)Source | 90.1% | — | Not comparable |
| BioMysteryBench (human-difficult)Source | 49.4% | — | Not comparable |
| SpatialBench VerifiedSource | 72.5% | — | Not comparable |
| SingleCellBenchSource | 60.6% | — | Not comparable |
| ProteinGym HardSource | 47.7% | — | Not comparable |
| Protein DesignSource | 42.5% | — | Not comparable |
| Organic chemistry V2Source | 61.6% | — | Not comparable |
| Protocols (troubleshooting)Source | 61.1% | — | Not comparable |
| Protocols (understanding)Source | 78.4% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 60.7% | 40.0% | Claude Opus 5 leads |
| AA-GPQA DiamondSource | 93.2% | 87.5% | Claude Opus 5 leads |
| AA-HLESource | 52.6% | 26.6% | Claude Opus 5 leads |
| AA-Omniscience IndexSource | 31.3% | -18.7% | Claude Opus 5 leads |
| AA-Omniscience AccuracySource | 54.2% | 37.5% | Claude Opus 5 leads |
| AA-Omniscience Hallucination RateSource | 50.1% | 89.8% | Claude Opus 5 leads |
| GPQASource | — | 88% | Not comparable |
Math7 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| IMO 2026Source | 42/42 | — | Not comparable |
| RiemannBench (no tools)Source | 60.0% | — | Not comparable |
| RiemannBench (tools)Source | 79.0% | — | Not comparable |
| ArXivMath Jun. 2026 (no tools)Source | 90.8% | — | Not comparable |
| ArXivMath Jun. 2026 (tools)Source | 91.3% | — | Not comparable |
| FrontierMath v2 (Tiers 1-3)Source | — | 28.280% | Not comparable |
| FrontierMath v2 (Tier 4)Source | — | 2.080% | Not comparable |
Multilingual3 benchmarks
MultimodalGPT-5.4 mini wins11 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| Chartography (no tools)Source | 29.6% | — | Not comparable |
| Chartography (tools)Source | 83.0% | — | Not comparable |
| BenchCAD Vision2Code (no tools)Source | 0.366 | — | Not comparable |
| BenchCAD Vision2Code (tools)Source | 0.821 | — | Not comparable |
| GDP.pdf (no tools)Source | 83.4% | — | Not comparable |
| GDP.pdf (tools)Source | 85.5% | — | Not comparable |
| OfficeQASource | 78.1% | — | Not comparable |
| OfficeQA ProSource | 66.9% | — | Not comparable |
| AA-MMMU-ProSource | 84.7% | 73.3% | Claude Opus 5 leads |
| MMMU-ProSource | — | 76.6% | Not comparable |
| MMMU-Pro w/ PythonSource | — | 78% | Not comparable |
Inst. Following1 benchmarks
| Benchmark | Claude Opus 5 | GPT-5.4 mini | Result |
|---|---|---|---|
| AA-IFBenchSource | — | 73.3% | Not comparable |
Frequently Asked Questions (4)
Which is better, Claude Opus 5 or GPT-5.4 mini?
Claude Opus 5 is ahead on BenchLM's BenchAlign leaderboard, 85.88 to 55.82. The biggest single separator in this matchup is HLE, where the scores are 64.7% and 41.5%.
Which is better for knowledge tasks, Claude Opus 5 or GPT-5.4 mini?
Claude Opus 5 has the edge for knowledge tasks in this comparison, averaging 64.7 versus 47.8. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, Claude Opus 5 or GPT-5.4 mini?
Claude Opus 5 has the edge for agentic tasks in this comparison, averaging 90.8 versus 65.7. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.
Which is better for multimodal and grounded tasks, Claude Opus 5 or GPT-5.4 mini?
GPT-5.4 mini has the edge for multimodal and grounded tasks in this comparison, averaging 76.6 versus 66.9. Inside this category, AA-MMMU-Pro is the benchmark that creates the most daylight between them.