Kimi K3.1 Pro vs Tencent Hy4: benchmarks, price and speed

This comparison puts Kimi's Kimi K3.1 Pro next to Tencent's Tencent Hy4. Their widest difference is the intelligence index, with Kimi K3.1 Pro ahead by 11.9 points (46.5 versus 34.6). What follows is built entirely from tracked data, so where the catalogue is silent, this page says so instead of guessing.

Overall capability: 11.9 points apart

The catalogue puts Kimi K3.1 Pro at 46.5 and Tencent Hy4 at 34.6 on its intelligence index. That is a wide gap for an aggregate score. In relative terms, Kimi K3.1 Pro scores 34% higher.

Across the 109 active models with an intelligence score in the catalogue, Kimi K3.1 Pro ranks 23rd and Tencent Hy4 ranks 42nd.

The catalogue does not yet split either model's score into coding and math indexes.

When they came out

Kimi K3.1 Pro came first, on October 23, 2026; Tencent Hy4 followed on November 3, 2026. They are effectively contemporaries.

What they cost

Kimi K3.1 Pro lists $2.00 per million input tokens and $10.00 per million output tokens; Tencent Hy4 lists $0.834 and $2.501. Kimi K3.1 Pro costs 4 times as much per output token.

A reference job of 1 million input tokens and 200,000 output tokens costs $4.00 with Kimi K3.1 Pro and $1.33 with Tencent Hy4, so Tencent Hy4 saves $2.67. A bigger example, a chat app with 50,000 sessions of 800 input and 300 output tokens, costs $230 on Kimi K3.1 Pro and $70.88 on Tencent Hy4: a gap of $159.

Score per dollar

A blunt value measure, blended price over intelligence score, puts Kimi K3.1 Pro at $0.086 and Tencent Hy4 at $0.036 per point. So Kimi K3.1 Pro buys its higher score at a premium, while Tencent Hy4 is the value pick, paying 58% less per point.

Tencent Hy4 is the faster writer

Measured output speed is 27 tokens per second for Kimi K3.1 Pro and 94 tokens per second for Tencent Hy4. On long outputs, that is a real advantage for Tencent Hy4. Time to first token is 3.0 seconds for Kimi K3.1 Pro and 2.4 seconds for Tencent Hy4. Kimi K3.1 Pro waits 1.3 times as long before its first token.

For a 600-token response, first token plus generation comes to roughly 25 seconds on Kimi K3.1 Pro and 8.8 seconds on Tencent Hy4.