Try Kimi at kimi.com →

Kimi K3.1 Pro Highspeed
Same Model, Faster Replies

K3.1 Pro Highspeed is the fast lane for Kimi's flagship model: the same K3.1 Pro answers, streamed about 2–3× faster by Moonshot's figure, for twice the price per token. Here's how fast it really is, what it costs and when it's worth it.

2–3× faster (Moonshot) $4 / $20 per 1M tokens 1M context
SAME PROMPT · SAME MODEL · DIFFERENT SPEED K K3.1 Pro Highspeedkimi-k3.1-pro-highspeed $4 in · $20 out ✓ Done · ~50 s K K3.1 Prokimi-k3.1-pro $2 in · $10 out ✓ Done · ~90 s Times from one independent spot check (AIHubMix). Prices per 1M tokens.
OVERVIEW

K3.1 Pro Highspeed at a glance

Launched with K3.1 Pro on 23 October 2026 in the Kimi apps, Kimi Work, Kimi Code and the Kimi API.

2–3×Faster output

Moonshot's figure. Independent checks measured about 1.7–2×.

$4 / $20Per 1M tokens

Input and output, double standard K3.1 Pro.

$0.20Cached input

The same as standard Pro, so cached prompts cost no more.

1MToken context

Same model, same abilities, same context window.

COMPARE

K3.1 Pro vs K3.1 Pro Highspeed

FeatureK3.1 ProK3.1 Pro Highspeed
API model IDkimi-k3.1-prokimi-k3.1-pro-highspeed
Kimi Code model IDk3.1-prok3.1-pro-highspeed
Model and abilitiesK3.1 ProSame model and abilities
Output speed (Moonshot)StandardAbout 2–3× faster
Output speed (independent checks)≈ 57–88 tokens/s≈ 113–150 tokens/s
Input, per 1M tokens$2.00$4.00
Cached input, per 1M tokens$0.20$0.20 (same)
Output, per 1M tokens$10.00$20.00
Cache write, 5 min / 1 hour$2.00 / $4.00$4.00 / $8.00
Context window1M tokens1M tokens
ThinkingAlways on: low, high or max effortSame
Kimi app accessAll plans (limited on Free)Pro plan ($39/month) and up
Kimi Code quotaStandardAbout 2× usage per task, Pro plan and up

Prices from the Kimi API pricing page, per million tokens, before tax. Full model details: Kimi K3.1 Pro guide.

SPEED

How much faster is it, really?

Moonshot says Highspeed runs at about 2–3× the speed of standard K3.1 Pro. Independent checks in October 2026 found a smaller but still large gain.

  • 1
    AIHubMix spot check: on one coding prompt, Highspeed streamed at 138–150 tokens/s against 85–88 for Pro, and finished in 43–60 seconds instead of 80–99.
  • 2
    Artificial Analysis data: endpoint figures reported by Kingy.ai put Pro at 57.5 tokens/s and Highspeed at 112.8, about twice as fast.
  • 3
    Why it matters: standard K3.1 Pro ranks among the slower top models, and it writes long reasoning before answering, so the extra speed is very noticeable in practice.

Speeds change with server load, prompt length and thinking effort. Treat these as a guide, not a guarantee.

MEASURED OUTPUT SPEEDTokens per second, higher is faster AIHubMix spot check ≈1.7× Pro 85–88 Highspeed 138–150 Artificial Analysis data ≈2.0× Pro 57.5 Highspeed 112.8 Moonshot says 2–3×. Independent checks so far: about 1.7–2×. Speeds change with load and prompt. Checks from October 2026. AA figures as reported by Kingy.ai from Artificial Analysis endpoint data.

Measured output speed from two independent checks.

CALCULATOR

Cost and speed calculator

Enter a typical request. See what Highspeed adds to your bill and how much waiting it saves.

K3.1 ProPer month–Output time per reply–
HighspeedPer month–Output time per reply–
DifferenceExtra cost per month–Waiting saved per month–

Assumes 70 tokens/s for standard Pro and excludes cache-write fees and thinking time before the first token. Estimates only.

WHEN

When to use Highspeed, and when not to

Use Highspeed when you're waiting

Interactive coding in Kimi Code, live chat features in your app, demos and pair-programming, where every reply holds you up.

Use Highspeed for long agent loops

Agents that make dozens of model calls in a row add up the wait each time. Faster output shortens the whole run.

Use standard Pro for batch work

Overnight jobs, bulk document processing and anything nobody watches. Speed doesn't matter, and Pro is half the price.

Use standard Pro on a tight budget

Highspeed doubles input and output prices. In Kimi Code it uses about twice the quota, so your plan runs out sooner.

Rule of thumb: if a person is waiting on the reply, Highspeed is usually worth it. If nobody is watching, use standard K3.1 Pro and save half the cost.
LIMITS

What Highspeed doesn't change

Only the model gets faster

Tool calls, web searches, test runs and scripts take the same time. If an agent spends most of its time running commands, the overall gain is smaller.

The first token still takes time

K3.1 Pro always thinks before it answers. Highspeed streams faster, but you may still wait a while before the first word appears. Lower the effort for quicker starts.

It isn't smarter

Highspeed has the same abilities as K3.1 Pro. Pick it for speed, not for better answers.

It's a hosted option

Highspeed is Moonshot's faster serving of K3.1 Pro. The open weights are for K3.1 Pro; speed when you self-host depends on your own hardware.

HOW

How to use K3.1 Pro Highspeed

Kimi app and Kimi Work

On the Pro plan or higher, pick K3.1 Pro Highspeed in the model menu. Kimi Work needs version 3.1.0 or later.

Kimi Code

On the Pro plan or higher, type /model and choose K3.1 Pro Highspeed. It uses about twice the quota.

Kimi API

Set the model to kimi-k3.1-pro-highspeed. Nothing else changes from standard K3.1 Pro.

import os from openai import OpenAI client = OpenAI( api_key=os.environ["MOONSHOT_API_KEY"], base_url="https://api.moonshot.ai/v1", ) stream = client.chat.completions.create( model="kimi-k3.1-pro-highspeed", messages=[{"role": "user", "content": "Refactor this function and explain the change: ..."}], reasoning_effort="high", # low | high | max (default max) stream=True, ) for chunk in stream: delta = chunk.choices[0].delta if delta.content: print(delta.content, end="", flush=True)

Keys, billing and rate limits: Kimi API guide. Coding setup: Kimi Code guide. Plan prices: Kimi pricing.

FAQ

K3.1 Pro Highspeed FAQ

Kimi K3.1 Pro Highspeed is a faster version of Kimi K3.1 Pro. It's the same model with the same abilities and 1M-token context, but Moonshot serves it so output arrives about 2–3 times faster. It costs twice as much per token.

Moonshot says about 2–3 times faster. Independent checks in October 2026 measured about 1.7–2 times faster: roughly 113–150 tokens per second against 57–88 for standard K3.1 Pro. Speeds vary with load and the prompt.

On the Kimi API it costs $4 per million input tokens and $20 per million output tokens, double standard K3.1 Pro. Cached input stays at $0.20 per million, and cache writes cost $4 (5 minutes) or $8 (1 hour) per million.

No. Moonshot says it has identical abilities. It gives the same quality of answers, only faster.

In the Kimi app and Kimi Code, Highspeed is available on the Pro plan ($39 a month) and above. It isn't on the Free or Plus plans. On the API, anyone with a funded account can use it.

Yes. Kimi Code's docs say Highspeed uses about twice the quota of standard K3.1 Pro for the same work.

Use kimi-k3.1-pro-highspeed on the Kimi API at https://api.moonshot.ai/v1. In Kimi Code, the model ID is k3.1-pro-highspeed.

Use Highspeed when someone is waiting on each reply, such as interactive coding or live chat. Use standard K3.1 Pro for batch jobs, background agents and cost-sensitive work, since it's half the price.

Try K3.1 Pro Highspeed

Switch the model ID in your API calls, or pick Highspeed in Kimi Code on the Pro plan.