GPT-5.6 Luna Max vs Qwen3.8 27B Xhigh: which is faster?

GPT-5.6 Luna Max leads the intelligence index 23.8 to 20.1. GPT-5.6 Luna Max costs 60% less on blended price. Compare price, speed, latency and a clear verdict.

Metric	GPT-5.6 Luna (Max)	Qwen3.8 27B (Xhigh)
Intelligence Index	23.8	20.1
Input price (per 1M tokens)	$0.20	$0.50
Output price (per 1M tokens)	$1.20	$3.00
Output speed (tokens/s)	125	38
Time to first token	108s	4.1s
Release date	2026-07-09	2026-08-14

Metric

GPT-5.6 Luna (Max)

Qwen3.8 27B (Xhigh)

Intelligence Index

23.8

20.1

Input price (per 1M tokens)

$0.20

$0.50

Output price (per 1M tokens)

$1.20

$3.00

Output speed (tokens/s)

125

Time to first token

108s

4.1s

Release date

2026-07-09

2026-08-14

This comparison puts OpenAI's GPT-5.6 Luna (Max) next to Alibaba's Qwen3.8 27B (Xhigh). The biggest gap is output speed, where GPT-5.6 Luna Max records 125 tokens per second and Qwen3.8 27B Xhigh 38 tokens per second. All numbers are the catalogue's recorded values; nobody ran either model by hand for this page.

Throughput and time to first token

Once a response is under way, GPT-5.6 Luna Max produces roughly 125 tokens per second against 38 tokens per second for Qwen3.8 27B Xhigh. That makes GPT-5.6 Luna Max roughly 3.3 times as fast at writing. Before any text appears, GPT-5.6 Luna Max takes 108 seconds and Qwen3.8 27B Xhigh 4.1 seconds. So the two pull in opposite directions, with Qwen3.8 27B Xhigh quicker off the mark and GPT-5.6 Luna Max quicker once it gets going; the crossover is around 5,800 tokens. A first-token wait of that length suits background work better than live conversation.

For a 1,000-token response, first token plus generation comes to roughly 116 seconds on GPT-5.6 Luna Max and 30 seconds on Qwen3.8 27B Xhigh.

Token pricing and a sample bill

The list prices are $0.20 in and $1.20 out for GPT-5.6 Luna Max, and $0.50 in and $3.00 out for Qwen3.8 27B Xhigh, per million tokens. Qwen3.8 27B Xhigh costs 2.5 times as much per output token.

Take 1 million input tokens and 200,000 output tokens: that comes to $0.44 on GPT-5.6 Luna Max and $1.10 on Qwen3.8 27B Xhigh. A bigger example, a chat app with 50,000 sessions of 800 input and 300 output tokens, costs $26.00 on GPT-5.6 Luna Max and $65.00 on Qwen3.8 27B Xhigh: a gap of $39.00.

The headline benchmark

GPT-5.6 Luna Max records 23.8 on the intelligence index and Qwen3.8 27B Xhigh records 20.1. It is a meaningful gap rather than a rounding difference. In relative terms, GPT-5.6 Luna Max scores 18% higher.

Across the 109 active models with an intelligence score in the catalogue, GPT-5.6 Luna Max ranks 70th and Qwen3.8 27B Xhigh ranks 78th.

There are no coding or math index scores for either model in the catalogue, which keeps this comparison at the aggregate level.

Score per dollar

Per point of intelligence index, GPT-5.6 Luna Max costs $0.019 of blended price per million tokens and Qwen3.8 27B Xhigh $0.056. With the higher score and the lower cost per point, GPT-5.6 Luna Max is the better value per point on these numbers.

Which is newer

GPT-5.6 Luna Max came first, on July 9, 2026; Qwen3.8 27B Xhigh followed on August 14, 2026. They are effectively contemporaries.

FAQ

Is GPT-5.6 Luna Max better than Qwen3.8 27B Xhigh?

It depends on what you need: GPT-5.6 Luna Max scores higher on the intelligence index, 23.8 to 20.1; GPT-5.6 Luna Max is 60% cheaper on blended price; GPT-5.6 Luna Max streams output 3.3 times as fast; and Qwen3.8 27B Xhigh reaches its first token sooner.

When were GPT-5.6 Luna Max and Qwen3.8 27B Xhigh released?

Qwen3.8 27B Xhigh is newer. It was released on August 14, 2026, 36 days after GPT-5.6 Luna Max (July 9, 2026).

Is Qwen3.8 27B Xhigh cheaper than GPT-5.6 Luna Max?

GPT-5.6 Luna Max is cheaper on a blended basis. GPT-5.6 Luna Max lists $0.20 per million input tokens and $1.20 per million output tokens; Qwen3.8 27B Xhigh lists $0.50 and $3.00. For 1 million input and 200,000 output tokens that is $0.44 against $1.10.

Where do these numbers come from?

The figures are the recorded benchmark, price and speed data in Franklin AI's model catalogue. This page is written from that data automatically; neither model was tested by hand for it.

Which is faster, GPT-5.6 Luna Max or Qwen3.8 27B Xhigh?

GPT-5.6 Luna Max generates output faster, at 125 tokens per second against 38 tokens per second. Qwen3.8 27B Xhigh reaches its first token sooner (4.1 seconds against 108 seconds). Counting the first-token wait, Qwen3.8 27B Xhigh finishes short replies first and GPT-5.6 Luna Max pulls ahead beyond about 5,800 tokens.

Final whistle

In one line: GPT-5.6 Luna Max has more going for it here, but Qwen3.8 27B Xhigh is not without an argument. Pick GPT-5.6 Luna Max if cost matters ($1.20 against $3.00 per million output tokens) and your outputs often run past about 5,800 tokens. Pick Qwen3.8 27B Xhigh if users are waiting on the first word and you want the more recent release.

Where the numbers come from

The scores, speed and prices in this comparison come from independent third-party model evaluations that Franklin AI syncs each day, and the article was written from a snapshot of them. For the latest results, open the GPT-5.6 Luna (Max) and Qwen3.8 27B (Xhigh) pages. The AI Benchmark methodology explains how that data is collected and shown.