On the Chompute AI Control Plane

Flat-rate Kimi K2.5. No token meter, no surprises.

Kimi K2.5 is an open-weight model tuned for tool-use and coding. Chompute runs it on a flat per-seat plan — not a per-token meter — so you can point agents and coding sessions at it for a cost you set once. Same Anthropic and OpenAI-compatible APIs, governed on the control plane.

Flat rate

Predictable cost, not a per-token meter.

The reason to run Kimi K2.5 here: one seat, one price, no usage math.

One line item

Flat $100 per seat, per month — one SKU, no usage reconciliation on finance reviews.

No token meter

Stop pricing every prompt. Point agents and coding sessions at it without a per-token bill climbing.

Paced, never cut off

Sustained heavy use eases off to stay within plan — predictable cost, no hard stops mid-run.

The model

What Kimi K2.5 is good at.

An open-weight model with a clear specialty: driving tool-using agents and writing code.

Tool-use & function calling

Built for agent loops: reliable tool calls, structured outputs, and multi-step plans that actually execute.

Coding

Generation, review, and multi-file refactors across real repositories.

Long-context work

Keeps large inputs in working memory — whole codebases, long documents, and full transcripts in one pass.

Open-weight efficiency

A Mixture-of-Experts design that delivers strong results well below frontier-closed prices.

Access

Run it under the control plane.

Kimi K2.5 runs under the Chompute AI Control Plane. Point your existing SDK at the gateway and call it over the Anthropic Messages and OpenAI-compatible APIs — streaming and tool-calling included, with the same keys, policy, and audit as every approved model.

How you reach it
  • Anthropic Messages, OpenAI Chat Completions, and OpenAI Responses
  • Drop-in over your existing SDK — no rewrite
  • Streaming and function/tool calling supported
  • Per-seat attribution, budgets, and audit on the control plane
Plan · Kimi K2.5
$100

Per seat · per month

Flat rate. No token meter.

Provisioned and governed under the Chompute AI Control Plane.

What’s included
  • Kimi K2.5 over Anthropic Messages, OpenAI Chat Completions, and OpenAI Responses
  • Flat per-seat pricing — no per-token meter
  • Open weights — inspectable, portable, no single-vendor lock-in
  • Tuned for tool-use, coding, and multi-step agents
  • Per-seat attribution and audit trail
  • Billed monthly; volume and annual terms on request
Get started

Run Kimi K2.5 on a flat-rate seat.

A short briefing and we’ll provision your team. No procurement cycle required.