GoogleGemini 2.5 ProVSOpenAIGPT-5.6 Luna

Analysis by:the whichllmmodel Editorial Team|Updated: June 2026

Our Take

These models use different coding benchmarks — with Gemini 2.5 Pro evaluated on swe-bench-verified (good at editing existing code, cross-file updates, and multi-component systems) and GPT-5.6 Luna on swe-bench-pro (excellent at multi-file repositories, autonomous agents, and industrial codebases) — but GPT-5.6 Luna holds a clear reasoning advantage (+7.9% on GPQA Diamond) while being cheaper or equal in cost to Gemini 2.5 Pro. We recommend going with GPT-5.6 Luna for superior overall value and reasoning capabilities.

Was this recommendation helpful?

Model Specs

Gemini 2.5 Pro

Website

Benchmarks & Scores

Coding (swe-bench-verified)

59.6%

good at editing existing code, cross-file updates, and multi-component systems

Reasoning (gpqa-diamond)

84.4%

graduate-level science QA

Cost & Context

Cost (per 1M tokens)

$3.44Input: $1.25 | Output: $10.00

Context Window

1.05M tokens

Model Specs

GPT-5.6 Luna

Website

Benchmarks & Scores

Coding (swe-bench-pro)

62.7%

excellent at multi-file repositories, autonomous agents, and industrial codebases

Reasoning (gpqa-diamond)Winner (+7.9%)

92.3%

graduate-level science QA

Cost & Context

Cost (per 1M tokens)1.5x cheaper

$2.25Input: $1.00 | Output: $6.00

Context WindowLarger

1.05M tokens

Read our data collection methodology

Frequently Asked Questions about Gemini 2.5 Pro vs GPT-5.6 Luna

GPT-5.6 Luna is cheaper than Gemini 2.5 Pro. GPT-5.6 Luna has a blended cost of $2.25/1M tokens, which is about 1.5x cheaper than Gemini 2.5 Pro at $3.44/1M tokens.

For coding tasks, Gemini 2.5 Pro scores 59.6% on swe-bench-verified (good at editing existing code, cross-file updates, and multi-component systems), while GPT-5.6 Luna scores 62.7% on swe-bench-pro (excellent at multi-file repositories, autonomous agents, and industrial codebases).

Related Matchups

Explore similar comparisons for Gemini 2.5 Pro and GPT-5.6 Luna.

Browse More Comparisons

Do you want to find a model for your constraints?

Use our interactive model finder to filter LLMs by reasoning capability, coding performance, cost, and context length.

Open Model Finder