· Updated · CostMyTokens · Analysis · 5 min read

Kimi 2.7 Pricing (Nov 9, 2026): $0.95/$4 Retiring · K3.1 Pro $2/$10

Kimi 2.7 pricing Nov 9, 2026: Kimi K2.7 Code $0.95/$4 per 1M (closed to new users, K2 retiring end of Dec) — current flagship K3.1 Pro $2/$10, K3 $3/$15, Batch 60%.

Updated Nov 9, 2026. Searching for Kimi 2.7 pricing? Kimi K2.7 Code is $0.95 input / $4.00 output per 1M tokens on Moonshot’s official list — but it is closed to new users since October 31, 2026, and Moonshot plans to retire the whole K2 series by the end of December 2026. The current flagship is Kimi K3.1 Pro at $2 / $10 per 1M (K3 $3 / $15). This page lists official Kimi API pricing for every model still on the price list, then shows what a month costs and where Kimi sits against DeepSeek and the closed flagships.

All prices are Moonshot’s official USD list prices from the Kimi API Platform pricing docs and the platform.kimi.ai homepage, re-checked 2026-11-02. CNY prices on platform.moonshot.cn and Kimi consumer memberships are separate products and are not mixed in. Standard workload for monthly figures: 10M tokens, split 8M input / 2M output, no caching.

Disclosure: this article contains no affiliate links. Where CostMyTokens does use affiliate links elsewhere, they never affect order — every table is sorted by price only. Read our disclosure.

Kimi API prices (official, per 1M tokens)

ModelInput / 1MOutput / 1MCache hit / 1MCache write 5m / 1hContextStatus
Kimi K3.1 Pro$2.00$10.00$0.20$2.00 / $4.001,048,576Current flagship
Kimi K3.1 Pro Highspeed$4.00$20.00$0.20$4.00 / $8.001,048,576Current (≈3× speed)
Kimi K3$3.00$15.00$0.30$3.00 / $6.001,048,576Previous flagship
Kimi K2.7 Code$0.95$4.00$0.19—262,144Closing (K2 series)
Kimi K2.7 Code Highspeed$1.90$8.00$0.38—262,144Closing (K2 series)
Kimi K2.6$0.95$4.00$0.16—262,144Closing (K2 series)

Source: Kimi model inference pricing, verified 2026-11-02. Pricing is flat across the full context window — there is no long-context surcharge on Kimi API pricing, unlike Grok’s 200K rule or Gemini’s 200K tier.

Kimi Batch API: 60% of standard

Moonshot’s Batch API bills at 60% of the standard price (a 40% discount) and supports kimi-k3.1-pro, kimi-k3, kimi-k2.7-code and kimi-k2.6. The official BatchJob pricing table lists these rates:

ModelBatch input / 1MBatch output / 1MBatch cache hit / 1M
Kimi K3.1 Pro$1.20$6.00$0.12
Kimi K3$1.80$9.00$0.18
Kimi K2.7 Code$0.57$2.40$0.114
Kimi K2.6$0.57$2.40$0.10

Batch jobs on K3.1 Pro and K3 are still charged cache-write fees; K2.7 Code and K2.6 do not support cache writes. Cache hits in batch mode are best-effort, not guaranteed. The Highspeed variants are not on the Batch list.

Kimi K2.7 pricing and retirement

Kimi K2.7 is sold on the API as Kimi K2.7 Code (model id kimi-k2.7-code), plus a Highspeed variant. Kimi K2.7 pricing is $0.95 input, $4.00 output and $0.19 cache hit per 1M tokens — the same base rates as K2.6, and less than half of K3.1 Pro.

The catch is the schedule. Moonshot’s official notice: starting October 31, 2026, K2 series models are no longer available to newly registered users (existing users are not affected), and Moonshot plans to retire all K2 series models by the end of December 2026. Moonshot has not published an exact shutdown day — only “end of December”. If you are on K2.7 Code today, budget a migration to K3.1 Pro before year end; new accounts cannot start on K2.7 at all.

What a month costs

10M tokens, 8M input / 2M output, standard tier, no caching. Sorted by cost.

ModelStandardBatch (60%)
Kimi K2.7 Code / K2.6$15.60$9.36
Kimi K3.1 Pro$36.00$21.60
Kimi K3$54.00$32.40
Kimi K3.1 Pro Highspeed$72.00—
Kimi K2.7 Code Highspeed$31.20—

Two things stand out. K3.1 Pro is cheaper than K3 on every column, so there is no price reason to stay on the previous flagship. And moving from K2.7 Code to K3.1 Pro costs about 2.3× at the standard mix — the real price of the K2 retirement.

Kimi against the rivals

At the same 8M/2M mix, Kimi K3.1 Pro ($36.00) lands level with the mainstream closed tier — Claude Sonnet 5.5 and GPT-6.1 Sol also come to $36 — while DeepSeek’s flagship is far cheaper: deepseek-v4-pro costs $18.48 at peak and $9.24 off-peak. Kimi’s advantages are the 1M context at a flat rate and a 40% Batch discount that applies to every token type. For the full price ranking, see cheapest LLM API in 2026.

FAQ

What is Kimi API pricing in 2026? Kimi API pricing for the current flagship, Kimi K3.1 Pro, is $2.00 input and $10.00 output per 1M tokens, with cache hits at $0.20. Kimi K3 costs $3.00 / $15.00 and K3.1 Pro Highspeed costs $4.00 / $20.00.

How much does Kimi K2.7 cost? Kimi K2.7 Code costs $0.95 input, $4.00 output and $0.19 cache hit per 1M tokens on the official list, and $0.57 / $2.40 on Batch. It has been closed to new users since October 31, 2026.

When is Kimi K2.7 retired? Moonshot plans to retire all K2 series models, including K2.7 Code and K2.6, by the end of December 2026. No exact day has been published; existing users keep access until then.

Does the Kimi API have a Batch discount? Yes. Batch inference costs 60% of the standard price for K3.1 Pro, K3, K2.7 Code and K2.6 — for example $1.20 / $6.00 per 1M on K3.1 Pro.

Is Kimi K3 pricing higher than K3.1 Pro? Yes. Kimi K3 pricing is $3.00 / $15.00 per 1M, while the newer K3.1 Pro is $2.00 / $10.00, so K3.1 Pro is the cheaper default.

Browse every official Kimi model on the Moonshot (Kimi) provider page, price your own traffic on the Kimi K3.1 Pro or Kimi K2.7 Code model pages, and compare with DeepSeek API pricing, Grok API pricing and the cheapest LLM API in 2026.

Prices: Kimi model inference pricing and Kimi BatchJob pricing, verified 2026-11-02. Prices change; check the official page before committing budget.

Share:
Back to Blog

Related Posts

View All Posts »