Models

kimi-k3

sference · 1049K ctx · 7 providers · updated 2026-08-04 22:15 UTC

Cheapest
$4.50 via Requesty
Avg price
$5.79 /1M
providers
7
Tokens (24h)
151B
Volume (24h)
$908.8K est.
(3 × $2.25 + $11.25) ÷ 4 = $4.50/1M — the $/1M figure blends input and output at 3:1 — that is the only way models with two different rates can be ranked on one axis. Your actual rate depends on your own input:output mix — input only $2.25, 3:1 $4.50, 1:1 $6.75, output only $11.25 per 1M.

The board shows this model's cheapest measured offer ($4.50/1M); the smaller figure under it is the mean of the 7 offers listed below ($5.79/1M). Add those up and divide — it matches. 24h token volume (151B) is sourced from OpenRouter public rankings — today's only public source of absolute per-model token counts (Vercel AI Gateway publishes share-of-traffic percentages); as more sources come online, the per-router breakdown will sum to this total here. Cross-check — Vercel AI Gateway: 4.32% of gateway token volume, 4.69% of spend (2026-08-04; share-of-traffic, daily, data CC BY 4.0).

ProviderRouter / Source$ in /1M$ out /1MTTFTTPSUptimeEvidence
requesty Requesty $2.25$11.25 API-measured
openrouter-best OpenRouter $3.00$15.00 API-measured
together HuggingFace Router $3.00$15.00 786ms 35 API-measured
fireworks-ai HuggingFace Router $3.00$15.00 919ms 41 API-measured
baseten HuggingFace Router $3.00$15.00 951ms 75 API-measured
0g-teetls 0G Router $3.00$15.00 API-measured
novita-ai Novita AI $3.00$15.00 API-measured

Volume trend

242B 18d d-17 · 72B tokensd-16 · 142B tokensd-15 · 163B tokensd-14 · 162B tokensd-13 · 190B tokensd-12 · 185B tokensd-11 · 186B tokensd-10 · 186B tokensd-9 · 158B tokensd-8 · 165B tokensd-7 · 185B tokensd-6 · 219B tokensd-5 · 242B tokensd-4 · 191B tokensd-3 · 217B tokensd-2 · 194B tokensd-1 · 175B tokensd-0 · 151B tokens

Last 18 days: 72B → 151B (+110%); range 72B–242B tokens/day. Source: OpenRouter public rankings; refreshed hourly.

Data & method

Cheapest ranked by blended = (3×in + out) ÷ 4; performance columns show only when a source provides them, never estimated. “Official list” = LiteLLM open pricebook (MIT); “API-measured” = provider/aggregator public APIs. Machine-readable data: /api/v1/models/kimi-k3.json. Free reuse requires attribution to tkx.org.

About

Kimi K3 is a machine-learning system that serves as a large language model, designed to process and generate human-like text based on input data. It is listed under the organization tag "moonshotai" and is available on AI model marketplaces for various inference providers to utilize.

AI-generated summary from public information — is this yours?

FAQ

How much does kimi-k3 cost?

Cheapest measured offer right now: $2.25/1M input, $11.25/1M output via requesty — across 7 tracked providers, refreshed hourly.

Who serves kimi-k3?

7 providers currently serve it. The table above lists each with input/output price, TTFT, TPS, uptime and an evidence tier.

How heavily is kimi-k3 used?

kimi-k3 routed 151B tokens in the last 24h. The daily bars above show routed volume over recent days. Source: OpenRouter public rankings — one of the few routers publishing per-model usage; TKX refreshes hourly.

Related analysis

Back to rankings