GPT-6 Luna vs Qwen3.7 Flash

The match-up at a glance

GPT-6 Luna and Qwen3.7 Flash are two budget models, built for high volumes at a low price, one from OpenAI, the other from Alibaba Cloud (Qwen). GPT-6 Luna was released on September 22, 2026, Qwen3.7 Flash on July 15, 2026.

All figures below come from public sources checked on October 22, 2026: models.dev and OpenRouter for prices and capabilities, Epoch AI for scores.

What the benchmarks say

On our capability index, GPT-6 Luna scores 80.3 out of 100 and Qwen3.7 Flash 63.7. GPT-6 Luna leads by 16.6 points.

On the 3 benchmarks reported for both, GPT-6 Luna wins 3 times and Qwen3.7 Flash 0:

• GPQA Diamond (PhD-level science questions): GPT-6 Luna 90.5%, Qwen3.7 Flash 82.3%
• OTIS Mock AIME (competition mathematics problems): GPT-6 Luna 98.9%, Qwen3.7 Flash 86.7%
• FrontierMath (very hard research-level mathematics): GPT-6 Luna 78.9%, Qwen3.7 Flash 19.3%

Which one costs less?

GPT-6 Luna costs $0.10 per million input tokens and $0.50 per million output tokens; Qwen3.7 Flash costs $0.03 and $0.13.

For a typical agent task (read 20,000 tokens, write 2,000), expect $0.003 with GPT-6 Luna versus $0.001 with Qwen3.7 Flash. Qwen3.7 Flash is 3.5 times cheaper, a gap that adds up quickly over 1,000 tasks: $2.14 difference.

Run the numbers with your own volumes in the agent cost calculator.

Features and limits

GPT-6 Luna reads more in one go: 1,050,000 tokens of context (roughly 1,575 pages), versus 1,000,000 for Qwen3.7 Flash.

Reasoning mode: GPT-6 Luna yes, Qwen3.7 Flash yes. Tool calling: GPT-6 Luna yes, Qwen3.7 Flash yes. Image input: GPT-6 Luna yes, Qwen3.7 Flash yes.

Neither publishes its weights: both are used through an API.

In short

Our reading of the data:

• Choose GPT-6 Luna for a higher capability index and a more recent release.
• Choose Qwen3.7 Flash for a lower cost.

This verdict relies on public figures; it does not replace a trial on your own tasks.