Which LLM is the cheapest for our workload?

The best-for pages rank the live catalogue per task by the index that measures it, and the router picks the cheapest model that meets your quality rule for each request.

How much can we save on our LLM bill?

On the public benchmark the router saved 89% against sending everything to the reference model. Import last month's provider bill on the bill page and the estimate is computed on your own traffic before you change a line of code.

How often do model prices change?

The catalogue is refreshed from the vendors' price lists; each model page shows the date of its data.

Model catalogue

Moonshot AI

Kimi K2.5

moonshotai/kimi-k2.5Catalogue: Indexed in the reference catalogue; nothing on this deployment answers as this model yet.

JSON

Specification

Published by the vendor; the platform bills at the target's declared price.

Kimi K2.5

reasoning open weights tools

Moonshot AI moonshotai/kimi-k2.5

Kimi K2.5 is Moonshot AI's native multimodal model, delivering state-of-the-art visual coding capability and a self-directed agent swarm paradigm. Built on Kimi K2 with continued pretraining over approximately 15T mixed...

Input $/1M: $0.50
Output $/1M: $2.50
Context: 262K
Max output: 236K
Released: Jan 27, 2026
Modalities: textimage

Intelligence index24

Coding index47

Agentic index-

Cached input reads $0.15 per 1M · weights on Hugging Face as moonshotai/Kimi-K2.5

Alternatives, and where the router puts Kimi K2.5

Kimi K2.5 earns a request when it needs step-by-step reasoning, tool calls, very long inputs; a short, general request would go to GLM 5.2 instead - $0.01 per million input tokens against $0.50 here; and the hardest requests, where the quality floor rules it out, to Claude Opus 5.5 (batch). Your rules decide the floor; the trace shows the comparison for every request.

Cheaper alternatives

Lower input price and an intelligence index within 5 points.

Better alternatives

A higher intelligence index, best first, with the price it comes at.

On this deployment

Nothing on this deployment answers as this model yet. Connect your own provider under Model providers, or add a target to your catalogue - the router picks it up without retraining.

Not available here yet

Signed in, add the model under Model providers; self-hosting, add a target with metadata.model set to this id.

targets.yamlyaml

# targets.yaml - declare a target that runs Kimi K2.5
targets:
  - id: kimi-k2-5
    kind: llm
    name: Kimi K2.5
    metadata:
      model: moonshotai/kimi-k2.5          # links the target to this catalogue entry
    capabilities:
      domains: [general]
      supports_tools: true
      context_window: 262144
    cost:
      usd_per_1k_tokens: 0.001500

Price among Moonshot AI models

Input list price per million tokens; Kimi K2.5 is the 3rd cheapest of 10 listed.

1
Kimi Latest
$0.22
2
Kimi K3
$0.23
3
Kimi K2.5
$0.50
4
Kimi K2.7 Code
$0.56
5
Kimi K2 0711
$0.57
6
Kimi K2 Thinking
$0.60
7
Kimi K2 0905
$0.60
8
Kimi K2.6
$0.65
9
Kimi K3.1 Pro
$2.00
10
Kimi K3 (batch)
$2.28

More from Moonshot AI

Other models by the same vendor in the catalogue.

List price is where the decision starts

Kimi K2.5 earns a request when it beats the field on your objective, not on the vendor's page.

The specification above is what Moonshot AI publishes. What the router uses is the price you actually pay here, the latency measured on this deployment and the outcomes people reported - and it compares those against every other model that could take the request. Price a workload on the estimator, or send one live in the playground.

Photo: RDNE Stock project

For the business

What it means for which models you pay for.

Catalogue of every vendor
The right model, not the famous one
Price, context, abilities and benchmark indexes side by side, with the model that fits each task named.
The rankings
89% less
Lower cost per request
On the 150-prompt benchmark the router spent 89% less than sending every prompt to the reference model, and the judged quality went up, not down.
Read the benchmark
Per million tokens, every model
Prices you can compare
The catalogue prints input and output prices and what three typical monthly workloads cost on each.
The model catalogue

Which LLM is the cheapest for our workload?: The best-for pages rank the live catalogue per task by the index that measures it, and the router picks the cheapest model that meets your quality rule for each request.
How much can we save on our LLM bill?: On the public benchmark the router saved 89% against sending everything to the reference model. Import last month's provider bill on the bill page and the estimate is computed on your own traffic before you change a line of code.
How often do model prices change?: The catalogue is refreshed from the vendors' price lists; each model page shows the date of its data.

On this deployment

Nothing on this deployment answers as this model yet. Connect your own provider under Model providers, or add a target to your catalogue - the router picks it up without retraining.

Not available here yet

Signed in, add the model under Model providers; self-hosting, add a target with metadata.model set to this id.

targets.yamlyaml

# targets.yaml - declare a target that runs Kimi K2.5
targets:
  - id: kimi-k2-5
    kind: llm
    name: Kimi K2.5
    metadata:
      model: moonshotai/kimi-k2.5          # links the target to this catalogue entry
    capabilities:
      domains: [general]
      supports_tools: true
      context_window: 262144
    cost:
      usd_per_1k_tokens: 0.001500

Kimi K2.5 earns a request when it beats the field on your objective, not on the vendor's page.

Specification

Kimi K2.5

Alternatives, and where the router puts Kimi K2.5

Cheaper alternatives

Better alternatives

On this deployment

Price among Moonshot AI models

More from Moonshot AI

Kimi K2.5 earns a request when it beats the field on your objective, not on the vendor's page.

What it means for which models you pay for.

The right model, not the famous one

Lower cost per request

Prices you can compare

Specification

Kimi K2.5

Alternatives, and where the router puts Kimi K2.5

Cheaper alternatives

Better alternatives

On this deployment

Price among Moonshot AI models

More from Moonshot AI

Kimi K2.5 earns a request when it beats the field on your objective, not on the vendor's page.

What it means for which models you pay for.

The right model, not the famous one

Lower cost per request

Prices you can compare