GPT-6 Luna (low)vsClaude Opus 4 8 Thinking

GPT-6 Luna (low) leads the overall Podium Score by 0.5 points (74.5 vs 74.0).Category wins: GPT-6 Luna (low) 0 — 0 Claude Opus 4 8 Thinking.Claude Opus 4 8 Thinking is the cheaper pick ($15/M vs $20/M per 1M output tokens). GPT-6 Luna (low) is faster (150 t/s vs 50 t/s).

GPT-6 Luna (low)Drag slider or hover to compareClaude Opus 4 8 Thinking
Claude Opus 4 8 Thinking
74.0
Coding: — · Speed: 50 t/s
Output Price: $15/M
GPT-6 Luna (low)
74.5
Coding: 55.0 · Speed: 150 t/s
Output Price: $20/M
Quick VerdictAggregated 2026 Benchmark Analysis

GPT-6 Luna (low) is rated higher overall with a Podium Score of 74.5 (vs 74.0 for Claude Opus 4 8 Thinking). Both models show closely matched capabilities across domain benchmarks.

Coding & Engineering
Comparable
Inference Speed
GPT-6 Luna (low) (150 t/s)
Cost Efficiency
Claude Opus 4 8 Thinking ($15/M)
API Price / 1M (in · out)
GPT-6 Luna (low) $5/M · $20/M — Claude Opus 4 8 T $3/M · $15/M

Choose GPT-6 Luna (low) for low-latency applications, and Claude Opus 4 8 Thinking for cost-sensitive workloads.

MetricGPT-6 Luna (low)Claude Opus 4 8 Thinking
Podium Score74.5 ▲74.0
Arena Elo——
Intelligence Index——
Coding55.0—
Math——
Reasoning80.0—
Agentic——
Knowledge45.0—
Multimodal85.0—
Long-context——
Output speed150 t/s ▲50 t/s
Output price$20/M$15/M ▲
Context window1000K ▲128K
0 / 4