Fireworks AI · automatic caching

Kimi K2.7 Code

Kimi K2.7 Code on Zumik: live pricing, context, and caching, routable by id or alias through one OpenAI-compatible endpoint.

Call it through Zumik About Fireworks AI

$0.95

Input / 1M tokens

$4.00

Output / 1M tokens

$0.19

Cache read · −80%

262K

Context window

Specifications

At a glance.

Provider	Fireworks AI
Family	kimi_k25
Released	2026-06
License	Open weights
Context window	262K tokens
Max output	262K tokens
Parameters	1029B
Modalities	text, image
Tool calling	Yes
Reasoning mode	Yes
Caching	automatic
Batch discount	No batch tier

Measured by Zumik

What reuse looks like here.

Not yet profiled

Pricing, context, and capabilities for Kimi K2.7 Code are live, but it is outside the flagship set Zumik benchmarks in depth, so measured reuse, capture, and warm TTFT are not shown yet. Run a workload estimate or route it by id to start collecting traces.

Reuse economics

What you actually pay once caching works.

At a typical 55% prefix reuse, a million input tokens on Kimi K2.7 Code effectively costs $0.53 instead of $0.95 - blending to roughly $1.40 with a 25% output share. There is no batch tier, so cost control here leans on caching and routing.

Estimate it for your workload

Best for

Route it directly by id, or let an alias pick it when it wins under policy.

Call it

Same OpenAI client, this model.

python

from openai import OpenAI

client = OpenAI(base_url="https://api.zumik.ai/v1", api_key="zk_live_...")

r = client.responses.create(
    model="kimi-k2p7-code",
    input="Draft a fix for the failing test.",
)
print(r.usage.input_tokens_cached)   # confirm reuse

Frequently asked

Kimi K2.7 Code, answered.

How much does Kimi K2.7 Code cost?

Kimi K2.7 Code is $0.95 per million input tokens and $4.00 per million output tokens through Zumik. Cache reads are $0.19 per million, a 80% discount on input.

What is Kimi K2.7 Code's context window?

Kimi K2.7 Code supports a 262K-token context window with up to 262K output tokens.

Does Kimi K2.7 Code support prompt caching?

Yes. Fireworks AI uses Automatic prompt caching (serverless and dedicated) caching. In the Zumik corpus, Kimi K2.7 Code shows a median cache capture of 81% on agent workloads.

Run Kimi K2.7 Code with reuse measured.

Point an OpenAI client at Zumik and see exactly how much of this model's input you are reusing.

Migration quickstart Back to catalog