Kimi K3 Low vs Qwen3.8 27B Xhigh: price, speed and benchmarks

Kimi K3 (Low) and Qwen3.8 27B (Xhigh) are close rivals from Kimi and Alibaba. Price separates them more than anything else; on a blended basis Kimi K3 Low costs 5.3 times as much as Qwen3.8 27B Xhigh. Everything below is read from the benchmark, price and speed data in Franklin AI's catalogue; nothing here comes from a hands-on test.

What they cost

The list prices are $3.00 in and $15.00 out for Kimi K3 Low, and $0.50 in and $3.00 out for Qwen3.8 27B Xhigh, per million tokens. For output tokens, Kimi K3 Low is 5 times the price of Qwen3.8 27B Xhigh.

A reference job of 1 million input tokens and 200,000 output tokens costs $6.00 with Kimi K3 Low and $1.10 with Qwen3.8 27B Xhigh, so Qwen3.8 27B Xhigh saves $4.90. A bigger example, a month of coding-assistant traffic at 30 million input and 3 million output tokens, costs $135 on Kimi K3 Low and $24.00 on Qwen3.8 27B Xhigh: a gap of $111. Blended at three input tokens per output token, the effective rates are $6.00 for Kimi K3 Low and $1.125 for Qwen3.8 27B Xhigh.

When they came out

Kimi K3 Low came first, on July 16, 2026; Qwen3.8 27B Xhigh followed on August 14, 2026. They are effectively contemporaries.

Qwen3.8 27B Xhigh leads on the intelligence index

Qwen3.8 27B Xhigh comes out ahead on the intelligence index, 20.1 to 17.4. It is a small margin for an index that averages many evaluations. In relative terms, Qwen3.8 27B Xhigh scores 16% higher.

For context, that places Kimi K3 Low 86th and Qwen3.8 27B Xhigh 78th out of 109 scored models still on offer.

Neither model has a separate coding or math index recorded yet, so those narrower comparisons will have to wait for more data.

Is the price in proportion?

A blunt value measure, blended price over intelligence score, puts Kimi K3 Low at $0.34 and Qwen3.8 27B Xhigh at $0.056 per point. With the higher score and the lower cost per point, Qwen3.8 27B Xhigh is the better value per point on these numbers.

Speed and latency

Measured output speed is 49 tokens per second for Kimi K3 Low and 38 tokens per second for Qwen3.8 27B Xhigh. Kimi K3 Low is ahead by 29% on throughput. Time to first token is 3.6 seconds for Kimi K3 Low and 4.1 seconds for Qwen3.8 27B Xhigh.

Add the two numbers up for a 600-token reply and Kimi K3 Low finishes first: 16 seconds for Kimi K3 Low, 20 seconds for Qwen3.8 27B Xhigh.