Model comparison
Claude Opus 4.8 vs GPT-5.6 Luna
Head-to-head evidence from 28 shared benchmark results across 5 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: Claude Opus 4.8 #5 (Supported); GPT-5.6 Luna #22 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. Claude Opus 4.8 and GPT-5.6 Luna share 28 comparable benchmark results. 5 of 8 categories are comparable. 25 results are unique to Claude Opus 4.8; 13 to GPT-5.6 Luna.
Updated July 23, 2026- Shared results
- 28
- Claude Opus 4.8 only
- 25
- GPT-5.6 Luna only
- 13
- Comparable categories
- 5 / 8
Pick Claude Opus 4.8 if you want the stronger benchmark profile. GPT-5.6 Luna 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 28 shared benchmark results across 5 evidence categories; 5 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 4.8 is clearly ahead on the BenchAlign aggregate, 78.34 to 67.17. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude Opus 4.8's sharpest advantage is in coding, where it averages 81.1 against 62.7. The single biggest benchmark swing on the page is FrontierMath v2 (Tiers 1-3), 47.241% to 78.600%. GPT-5.6 Luna does hit back in knowledge, so the answer changes if that is the part of the workload you care about most.
Claude Opus 4.8 is also the more expensive model on tokens at $5.00 input / $25.00 output per 1M tokens, versus $1.00 input / $6.00 output per 1M tokens for GPT-5.6 Luna. That is roughly 4.2x 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 4.8 | Δ | GPT-5.6 Luna |
|---|---|---|---|
| Knowledge | Claude Opus 4.862.7 | Margin→ 29.6 | GPT-5.6 Luna92.3 |
| Math | Claude Opus 4.853.9 | Margin→ 19.7 | GPT-5.6 Luna73.6 |
| Coding | Claude Opus 4.881.1 | Margin← 18.4 | GPT-5.6 Luna62.7 |
| Agentic | Claude Opus 4.880.3 | Margin→ 3.8 | GPT-5.6 Luna84.1 |
| Multimodal | Claude Opus 4.877.0 | Margin→ 1.4 | GPT-5.6 Luna78.4 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
FrontierMath v2 (Tiers 1-3)
MathA 47.241%B 78.600%Winner: GPT-5.6 LunaΔ 31.4FrontierMath v2 (Tiers 1-3): Claude Opus 4.8 scored 47.241%; GPT-5.6 Luna scored 78.600%. GPT-5.6 Luna wins this benchmark. - Source ↗
FrontierMath v2 (Tier 4)
MathA 31.250%B 58.500%Winner: GPT-5.6 LunaΔ 27.3FrontierMath v2 (Tier 4): Claude Opus 4.8 scored 31.250%; GPT-5.6 Luna scored 58.500%. GPT-5.6 Luna wins this benchmark. - Source ↗
Terminal-Bench 2.0
AgenticA 74.6%B 84.7%Winner: GPT-5.6 LunaΔ 10.1Terminal-Bench 2.0: Claude Opus 4.8 scored 74.6%; GPT-5.6 Luna scored 84.7%. GPT-5.6 Luna wins this benchmark. - Source ↗
SWE-bench Pro
CodingA 69.2%B 62.7%Winner: Claude Opus 4.8Δ 6.5SWE-bench Pro: Claude Opus 4.8 scored 69.2%; GPT-5.6 Luna scored 62.7%. Claude Opus 4.8 wins this benchmark. - Source ↗
GPQA
KnowledgeA 93.6%B 92.3%Winner: Claude Opus 4.8Δ 1.3GPQA: Claude Opus 4.8 scored 93.6%; GPT-5.6 Luna scored 92.3%. Claude Opus 4.8 wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | Claude Opus 4.8 | GPT-5.6 Luna | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | Claude Opus 4.8$5 input / $25 output | GPT-5.6 Luna$1 input / $6 output | GPT-5.6 Luna has the lower combined listed price. |
| Generation speedtokens per second | Claude Opus 4.8Not available | GPT-5.6 LunaNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | Claude Opus 4.8Not available | GPT-5.6 LunaNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | Claude Opus 4.81M | GPT-5.6 Luna1M | Listed context windows are equal. |
Benchmark Deep Dive
AgenticGPT-5.6 Luna wins25 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 74.6% | 84.7% | GPT-5.6 Luna leads |
| BrowseCompSource | 84.3% | 83.3% | Claude Opus 4.8 leads |
| DeepSearchQASource | 93.1% | — | Not comparable |
| OSWorld-VerifiedSource | 83.4% | — | Not comparable |
| Finance Agent v2Source | 53.9% | — | Not comparable |
| GDPval-AASource | 1594 | 1584 | Claude Opus 4.8 leads |
| MCP AtlasSource | 82.2% | — | Not comparable |
| ToolathlonSource | 59.9% | 53.4% | Claude Opus 4.8 leads |
| Gert LabsSource | 72.97% | — | Not comparable |
| AA Agentic IndexSource | 47.2% | 45.6% | Claude Opus 4.8 leads |
| τ²-bench resultsSource | 94.4% | — | Not comparable |
| GDPval-AASource | 54.7% | 54.2% | Claude Opus 4.8 leads |
| ResearchClawBenchSource | 21.1% | — | Not comparable |
| OSWorld 2.0Source | 20.6% | 45.6% | GPT-5.6 Luna leads |
| AA BriefcaseSource | 1347 | — | Not comparable |
| AA AutomationBenchSource | 48.5% | 42.2% | Claude Opus 4.8 leads |
| AA EnterpriseOps-GymSource | 44.0% | — | Not comparable |
| AA Harvey LABSource | 91.1% | 87.9% | Claude Opus 4.8 leads |
| AA Tau3 BankingSource | 27.6% | 27.2% | Claude Opus 4.8 leads |
| terminalBenchHardSource | 58.3% | — | Not comparable |
| aaTerminalBench21Source | 84.6% | 80.9% | Claude Opus 4.8 leads |
| CyberGymSource | — | 77.9% | Not comparable |
| ExploitGymSource | — | 12.4% | Not comparable |
| AA ITBenchSource | — | 40.3% | Not comparable |
| APEX-Agents-AASource | — | 35.8% | Not comparable |
CodingClaude Opus 4.8 wins12 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 88.6% | — | Not comparable |
| SWE-bench ProSource | 69.2% | 62.7% | Claude Opus 4.8 leads |
| SWE MultilingualSource | 84.4% | — | Not comparable |
| SWE MultimodalSource | 38.4% | — | Not comparable |
| Terminal-Bench 2.0Source | 74.6% | 84.7% | GPT-5.6 Luna leads |
| cursorBench31Source | 58.4% | — | Not comparable |
| cursorBench32Source | 62.3% | 61.1% | Claude Opus 4.8 leads |
| AA Coding IndexSource | 74.3% | 71.5% | Claude Opus 4.8 leads |
| AA-SciCodeSource | 53.5% | 52.5% | Claude Opus 4.8 leads |
| FrontierCode 1.1 MainSource | 46.5% | — | Not comparable |
| deepSweSource | — | 67.2% | Not comparable |
| FrontierCode 1.1 ExtendedSource | — | 55.1% | Not comparable |
Reasoning3 benchmarks
KnowledgeGPT-5.6 Luna wins12 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| GPQASource | 93.6% | 92.3% | Claude Opus 4.8 leads |
| GPQA-DSource | 93.6% | 92.3% | Claude Opus 4.8 leads |
| HLESource | 57.9% | — | Not comparable |
| HLE w/o toolsSource | 49.8% | — | Not comparable |
| Artificial Analysis Intelligence IndexSource | 55.7% | 51.2% | Claude Opus 4.8 leads |
| AA-GPQA DiamondSource | 92.0% | 91.1% | Claude Opus 4.8 leads |
| AA-HLESource | 45.7% | 37.2% | Claude Opus 4.8 leads |
| AA-Omniscience IndexSource | 27.4% | -11.2% | Claude Opus 4.8 leads |
| AA-Omniscience AccuracySource | 46.6% | 41.5% | Claude Opus 4.8 leads |
| AA-Omniscience Hallucination RateSource | 35.9% | 90.1% | Claude Opus 4.8 leads |
| HealthBench ProfessionalSource | — | 55.7% | Not comparable |
| HealthBench HardSource | — | 32.0% | Not comparable |
MathGPT-5.6 Luna wins4 benchmarks
Multilingual1 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| INCLUDESource | 87.6% | — | Not comparable |
MultimodalGPT-5.6 Luna wins8 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| OfficeQA ProSource | 66.2% | — | Not comparable |
| ScreenSpot ProSource | 87.9% | — | Not comparable |
| CharXivSource | 89.9% | — | Not comparable |
| CharXiv w/o toolsSource | 80.5% | — | Not comparable |
| Design Arena WebsiteSource | 1270 | — | Not comparable |
| MMMU-ProSource | — | 78.4% | Not comparable |
| MMMU-Pro w/ PythonSource | — | 79.5% | Not comparable |
| AA-MMMU-ProSource | — | 78.6% | Not comparable |
Inst. Following1 benchmarks
| Benchmark | Claude Opus 4.8 | GPT-5.6 Luna | Result |
|---|---|---|---|
| AA-IFBenchSource | 62.2% | — | Not comparable |
Frequently Asked Questions (6)
Which is better, Claude Opus 4.8 or GPT-5.6 Luna?
Claude Opus 4.8 is ahead on BenchLM's BenchAlign leaderboard, 78.34 to 67.17. The biggest single separator in this matchup is FrontierMath v2 (Tiers 1-3), where the scores are 47.241% and 78.600%.
Which is better for knowledge tasks, Claude Opus 4.8 or GPT-5.6 Luna?
GPT-5.6 Luna has the edge for knowledge tasks in this comparison, averaging 92.3 versus 62.7. Inside this category, AA-Omniscience Hallucination Rate is the benchmark that creates the most daylight between them.
Which is better for coding, Claude Opus 4.8 or GPT-5.6 Luna?
Claude Opus 4.8 has the edge for coding in this comparison, averaging 81.1 versus 62.7. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Which is better for math, Claude Opus 4.8 or GPT-5.6 Luna?
GPT-5.6 Luna has the edge for math in this comparison, averaging 73.6 versus 53.9. Inside this category, FrontierMath v2 (Tiers 1-3) is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, Claude Opus 4.8 or GPT-5.6 Luna?
GPT-5.6 Luna has the edge for agentic tasks in this comparison, averaging 84.1 versus 80.3. Inside this category, OSWorld 2.0 is the benchmark that creates the most daylight between them.
Which is better for multimodal and grounded tasks, Claude Opus 4.8 or GPT-5.6 Luna?
GPT-5.6 Luna has the edge for multimodal and grounded tasks in this comparison, averaging 78.4 versus 77. Claude Opus 4.8 stays close enough that the answer can still flip depending on your workload.
Related Comparisons
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.