Skip to main content

Model comparison

GPT-5.5 vs GPT-5.6 Luna

Data verified

Head-to-head evidence from 36 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.

OpenAI
73.51/100
Margin
6.3pts
← winning
67.17/100
0 category wins5 category wins

Public leaderboard positions: GPT-5.5 #9 (Estimated); GPT-5.6 Luna #22 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.

Evidence parity. GPT-5.5 and GPT-5.6 Luna share 36 comparable benchmark results. 5 of 8 categories are comparable. 21 results are unique to GPT-5.5; 5 to GPT-5.6 Luna.

Updated July 23, 2026
Shared results
36
GPT-5.5 only
21
GPT-5.6 Luna only
5
Comparable categories
5 / 8

Pick GPT-5.5 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 36 shared benchmark results across 6 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

GPT-5.5 is clearly ahead on the BenchAlign aggregate, 73.51 to 67.17. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

GPT-5.5 is also the more expensive model on tokens at $5.00 input / $30.00 output per 1M tokens, versus $1.00 input / $6.00 output per 1M tokens for GPT-5.6 Luna. That is roughly 5.0x 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 scores and score margins for GPT-5.5 and GPT-5.6 Luna
CategoryGPT-5.5ΔGPT-5.6 Luna
KnowledgeGPT-5.557.8Margin 34.5GPT-5.6 Luna92.3
MathGPT-5.547.6Margin 26.0GPT-5.6 Luna73.6
MultimodalGPT-5.570.4Margin 8.0GPT-5.6 Luna78.4
CodingGPT-5.558.6Margin 4.1GPT-5.6 Luna62.7
AgenticGPT-5.581.6Margin 2.5GPT-5.6 Luna84.1
ReasoningGPT-5.585.0MarginNo overlapGPT-5.6 LunaNot measured

Decisive benchmark drivers

The largest measured benchmark gaps in this matchup, with exact reported values.

More
A · GPT-5.5B · GPT-5.6 Luna
  1. FrontierMath v2 (Tiers 1-3)

    Math
    Source ↗
    A 51.700%B 78.600%
    Winner: GPT-5.6 LunaΔ 26.9
    FrontierMath v2 (Tiers 1-3): GPT-5.5 scored 51.700%; GPT-5.6 Luna scored 78.600%. GPT-5.6 Luna wins this benchmark.
  2. FrontierMath v2 (Tier 4)

    Math
    Source ↗
    A 35.400%B 58.500%
    Winner: GPT-5.6 LunaΔ 23.1
    FrontierMath v2 (Tier 4): GPT-5.5 scored 35.400%; GPT-5.6 Luna scored 58.500%. GPT-5.6 Luna wins this benchmark.
  3. SWE-bench Pro

    Coding
    Source ↗
    A 58.6%B 62.7%
    Winner: GPT-5.6 LunaΔ 4.1
    SWE-bench Pro: GPT-5.5 scored 58.6%; GPT-5.6 Luna scored 62.7%. GPT-5.6 Luna wins this benchmark.
  4. MMMU-Pro

    Multimodal
    Source ↗
    A 81.2%B 78.4%
    Winner: GPT-5.5Δ 2.8
    MMMU-Pro: GPT-5.5 scored 81.2%; GPT-5.6 Luna scored 78.4%. GPT-5.5 wins this benchmark.
  5. Terminal-Bench 2.0

    Agentic
    Source ↗
    A 82%B 84.7%
    Winner: GPT-5.6 LunaΔ 2.7
    Terminal-Bench 2.0: GPT-5.5 scored 82%; GPT-5.6 Luna scored 84.7%. GPT-5.6 Luna wins this benchmark.

Operational comparison

Runtime and commercial metrics are compared only when both models have a complete sourced value.

MetricGPT-5.5GPT-5.6 LunaComparison
Input / output priceUSD per 1M tokensGPT-5.5$5 input / $30 outputGPT-5.6 Luna$1 input / $6 outputGPT-5.6 Luna has the lower combined listed price.
Generation speedtokens per secondGPT-5.5Not availableGPT-5.6 LunaNot availableA complete speed comparison is not available.
First-answer latencyseconds to first tokenGPT-5.5Not availableGPT-5.6 LunaNot availableA complete latency comparison is not available.
Context windowmaximum listed tokensGPT-5.51MGPT-5.6 Luna1MListed context windows are equal.

Benchmark Deep Dive

AgenticGPT-5.6 Luna wins
BenchmarkGPT-5.5GPT-5.6 LunaResult
Terminal-Bench 2.0Source 82%84.7%GPT-5.6 Luna leads
CyberGymSource 81.8%77.9%GPT-5.5 leads
BrowseCompSource 84.4%83.3%GPT-5.5 leads
OSWorld-VerifiedSource 78.7%Not comparable
MCP AtlasSource 75.3%Not comparable
ToolathlonSource 55.6%53.4%GPT-5.5 leads
τ²-bench resultsSource 93.9%Not comparable
AA Agentic IndexSource 44.9%45.6%GPT-5.6 Luna leads
APEX-Agents-AASource 37.7%35.8%GPT-5.5 leads
GDPval-AASource 49.5%54.2%GPT-5.6 Luna leads
GDPval-AASource 14901584GPT-5.6 Luna leads
Gert LabsSource 72.93%Not comparable
ResearchClawBenchSource 17.0%Not comparable
OSWorld 2.0Source 13.0%45.6%GPT-5.6 Luna leads
JobBenchSource 42.7%Not comparable
ExploitGymSource 13.4%12.4%GPT-5.5 leads
AA BriefcaseSource 1154Not comparable
AA AutomationBenchSource 42.1%42.2%GPT-5.6 Luna leads
AA EnterpriseOps-GymSource 46.6%Not comparable
AA Harvey LABSource 86.3%87.9%GPT-5.6 Luna leads
AA ITBenchSource 45.8%40.3%GPT-5.5 leads
AA Tau3 BankingSource 31.3%27.2%GPT-5.5 leads
terminalBenchHardSource 60.6%Not comparable
aaTerminalBench21Source 84.3%80.9%GPT-5.5 leads
CodingGPT-5.6 Luna wins
BenchmarkGPT-5.5GPT-5.6 LunaResult
SWE-bench ProSource 58.6%62.7%GPT-5.6 Luna leads
Terminal-Bench 2.0Source 82.0%84.7%GPT-5.6 Luna leads
Vibe Code BenchSource 69.85%Not comparable
React Native EvalsSource 84.7%Not comparable
cursorBench31Source 59.2%Not comparable
cursorBench32Source 58.4%61.1%GPT-5.6 Luna leads
AA Coding IndexSource 74.9%71.5%GPT-5.5 leads
AA-SciCodeSource 56.1%52.5%GPT-5.5 leads
FrontierCode 1.1 MainSource 43.0%Not comparable
deepSweSource 67.2%Not comparable
FrontierCode 1.1 ExtendedSource 55.1%Not comparable
Reasoning
BenchmarkGPT-5.5GPT-5.6 LunaResult
MRCR v2 64K-128KSource 83.1%Not comparable
MRCR v2 128K-256KSource 87.5%Not comparable
ARC-AGI-2Source 85%Not comparable
AA-LCRSource 74.3%74.0%GPT-5.5 leads
CritPtSource 27.1%20.6%GPT-5.5 leads
ARC-AGI-3Source 0.2%Not comparable
KnowledgeGPT-5.6 Luna wins
BenchmarkGPT-5.5GPT-5.6 LunaResult
GPQASource 93.6%92.3%GPT-5.5 leads
GPQA-DSource 93.6%92.3%GPT-5.5 leads
HLESource 52.2%Not comparable
HLE w/o toolsSource 41.4%Not comparable
Artificial Analysis Intelligence IndexSource 54.8%51.2%GPT-5.5 leads
AA-GPQA DiamondSource 93.5%91.1%GPT-5.5 leads
AA-HLESource 44.3%37.2%GPT-5.5 leads
AA-Omniscience IndexSource 20.1%-11.2%GPT-5.5 leads
AA-Omniscience AccuracySource 56.9%41.5%GPT-5.5 leads
AA-Omniscience Hallucination RateSource 85.5%90.1%GPT-5.5 leads
HealthBench ProfessionalSource 55.7%Not comparable
HealthBench HardSource 32.0%Not comparable
MathGPT-5.6 Luna wins
BenchmarkGPT-5.5GPT-5.6 LunaResult
FrontierMath (legacy)Source 51.7%78.6%GPT-5.6 Luna leads
FrontierMath v2 (Tiers 1-3)Source 51.700%78.600%GPT-5.6 Luna leads
FrontierMath v2 (Tier 4)Source 35.400%58.500%GPT-5.6 Luna leads
MultimodalGPT-5.6 Luna wins
BenchmarkGPT-5.5GPT-5.6 LunaResult
MMMU-ProSource 81.2%78.4%GPT-5.5 leads
MMMU-Pro w/ PythonSource 83.2%79.5%GPT-5.5 leads
OfficeQA ProSource 54.1%Not comparable
AA-MMMU-ProSource 79.9%78.6%GPT-5.5 leads
Design Arena WebsiteSource 1282Not comparable
Inst. Following
BenchmarkGPT-5.5GPT-5.6 LunaResult
AA-IFBenchSource 75.9%Not comparable
Frequently Asked Questions (6)

Which is better, GPT-5.5 or GPT-5.6 Luna?

GPT-5.5 is ahead on BenchLM's BenchAlign leaderboard, 73.51 to 67.17. The biggest single separator in this matchup is FrontierMath v2 (Tiers 1-3), where the scores are 51.700% and 78.600%.

Which is better for knowledge tasks, GPT-5.5 or GPT-5.6 Luna?

GPT-5.6 Luna has the edge for knowledge tasks in this comparison, averaging 92.3 versus 57.8. Inside this category, AA-Omniscience Index is the benchmark that creates the most daylight between them.

Which is better for coding, GPT-5.5 or GPT-5.6 Luna?

GPT-5.6 Luna has the edge for coding in this comparison, averaging 62.7 versus 58.6. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.

Which is better for math, GPT-5.5 or GPT-5.6 Luna?

GPT-5.6 Luna has the edge for math in this comparison, averaging 73.6 versus 47.6. Inside this category, FrontierMath (legacy) is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, GPT-5.5 or GPT-5.6 Luna?

GPT-5.6 Luna has the edge for agentic tasks in this comparison, averaging 84.1 versus 81.6. Inside this category, GDPval-AA is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, GPT-5.5 or GPT-5.6 Luna?

GPT-5.6 Luna has the edge for multimodal and grounded tasks in this comparison, averaging 78.4 versus 70.4. Inside this category, MMMU-Pro w/ Python is the benchmark that creates the most daylight between them.

Related Comparisons

Last updated: July 23, 2026

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.