LLM Comparison

GPT-5.5 vs GPT-5.6 Luna: benchmark scores, pricing & comparison.

Side-by-side GPT-5.5 vs GPT-5.6 Luna comparison across SWE-bench, GPQA, HLE, Terminal-Bench, coding agent scores, token pricing, context window, and AskClash RWT. Green marks the winner on each benchmark.

Rank #7 vs #13AskClash overall scores 76.9 vs 67.9.
Pricing $5.00/$30.0 vs $1.00/$6.00Input and output token prices per 1M tokens when published.
Proprietary vs ProprietaryOpenAI vs OpenAI.

GPT-5.5 vs GPT-5.6 Luna benchmark comparison

Green cells highlight the winning model for each metric. Scores are cached from the AskClash LLM leaderboard snapshot.

MetricGPT-5.5GPT-5.6 Luna
Overall Score76.967.9
Leaderboard Rank#7#13
RWT8.09.5
Coding Agent Index76.475.0
HLE52.2
GPQA93.692.3
SWE-Pro58.662.7
SWE-Atlas81.0
Terminal-Bench82.784.7
DeepSWE67.067.2
OSWorld78.7
MCP Atlas75.3
Finance Agent51.855.0
MMMU-Pro81.278.4
ARC-AGI 285.059.5
Tau298.0
MRCR41.3
Input Price (per 1M tokens)$5.00$1.00
Output Price (per 1M tokens)$30.0$6.00
Context Window1M1M
Benchmarks Published1312

GPT-5.5 vs GPT-5.6 Luna head-to-head charts

GPT-5.5 leads 5 and GPT-5.6 Luna leads 5 of 10 shared benchmarks. GPT-5.6 Luna is cheaper on a defined 1M-input / 200K-output workload. Charts show only benchmarks both models publish.

GPT-5.5GPT-5.6 Luna
Overall
76.9GPT-5.5
67.9GPT-5.6 Luna
RWT
8.0GPT-5.5
9.5GPT-5.6 Luna
Coding Agent Index
76.4GPT-5.5
75.0GPT-5.6 Luna
GPQA
93.6GPT-5.5
92.3GPT-5.6 Luna
SWE-Pro
58.6GPT-5.5
62.7GPT-5.6 Luna
Terminal-Bench
82.7GPT-5.5
84.7GPT-5.6 Luna
DeepSWE
67.0GPT-5.5
67.2GPT-5.6 Luna
Finance Agent
51.8GPT-5.5
55.0GPT-5.6 Luna
MMMU-Pro
81.2GPT-5.5
78.4GPT-5.6 Luna
ARC-AGI 2
85.0GPT-5.5
59.5GPT-5.6 Luna
GPT-5.5
Input$5.00
Output$30.0
Workload$11
Context1M
GPT-5.6 Luna
Input$1.00
Output$6.00
Workload$2.20
Context1M

Workload = published cost of 1M input + 200K output tokens. Open the live leaderboard for interactive compare charts.

More GPT-5.5 and GPT-5.6 Luna comparisons

Explore how GPT-5.5 and GPT-5.6 Luna stack up against other top-ranked LLMs.

How to read this comparison

Benchmark scores

Higher is better for all benchmark scores (SWE-bench, GPQA, HLE, Terminal-Bench, etc.). Green marks the model with the higher score.

Token pricing

Lower is better for input and output prices. Green marks the cheaper model per 1M tokens.

Coverage matters

Models with fewer disclosed benchmark cells may have inflated percentile scores. Check the benchmark cell count for context.

This comparison page is generated from the AskClash LLM leaderboard cache. Open the live leaderboard for real-time scores and interactive filtering.