minirouter.sh
Open navigation
Get a key

Moonshot AI

Kimi K3.1 Pro

Moonshot's Kimi K3 upgrade for coding and multi-step work, with a million-token context window and adjustable reasoning effort.

Get a MiniRouter keySee the evidence
moonshotai/kimi-k3.1-protext inputtext output

Replaces Kimi K3: input $3.15 → $2.1 per 1M (-33%), output $15.75 → $10.5 per 1M (-33%), context 1M → 1M, max output 131K → 1M.

Run this model

moonshotai/kimi-k3.1-pro

Use an endpoint supported by this model.

Get a key for this setup →
Thinking guide →
const response = await fetch('https://api.minirouter.sh/v1/chat/completions', {  method: 'POST',  headers: {    Authorization: `Bearer ${process.env.MINIROUTER_KEY}`,    'Content-Type': 'application/json',  },  body: JSON.stringify({    model: 'moonshotai/kimi-k3.1-pro',    messages: [{ role: 'user', content: 'Why is the sky blue?' }],    max_tokens: 1024,  }),}) const data = await response.json()console.log(data.choices[0].message.content)
Read docs →

Start with a bounded task and an explicit max_tokens limit. Reasoning and the final answer share the output budget; use low effort when a shorter response matters.

Model overview

Input / 1M
$2.1
All-in MiniRouter rate
Output / 1M
$10.5
All-in MiniRouter rate
Context
1M
1,048,576 tokens
Max output
1M
tokens
Reasoning
Low · high · max
Always enabled
Predecessor
Kimi K3
Further post-training
Announced
23 Oct 2026
publisher release
Evaluate on your tasks

Give it a task with a checkable result.

Moonshot positions K3.1 Pro for longer coding tasks, scripted 3D work and research across documents. Test the completed artifact, including the steps your application executes.
Read the release analysis

A practical evaluation

Code changes
Run the testsCheck correctness across the affected project.
Research
Verify the sourcesTrace calculations and claims to the input documents.
Workload focus from the Moonshot announcement ↗. Evaluation suggestions are MiniRouter's.
View chart data
Suggested Kimi K3.1 Pro workload checks
TaskProvideVerify
Repository changeRelevant code, constraints and test commandsTests, diff and unintended changes
Blender or CAD scriptA brief and the target application's APIRun the script and inspect the geometry
Document researchSource documents and an explicit questionCitations, calculations and unsupported claims
Control the work

Budget for reasoning. Supply the tools.

K3.1 Pro always reasons. Choose low, high or max effort for the task, then measure completed results, latency and total cost against your existing model.
Publisher API guide
  1. [ 01 ]

    Start small

    Try a focused prompt at low effort before expanding to a long agent run. Leave enough output tokens for the answer after reasoning.

  2. [ 02 ]

    Keep the conversation intact

    When returning tool results, preserve the complete assistant message, including reasoning and tool calls.

  3. [ 03 ]

    Run tools in your application

    A tool call is a request for your application to act. Shells, Blender, CAD and document retrieval need an execution environment you provide.

Migration and access

Check the workflow before switching models.

The larger post-training run builds on Kimi K3. Moonshot's benchmark gains are useful starting evidence; your prompts, tools and acceptance checks determine whether it fits your application.
Moonshot announcement
  1. [ 01 ]

    Validate generated artifacts

    Run code and inspect outputs before using them. For research, check every material claim and calculation against its source.

  2. [ 02 ]

    Use the advertised route capabilities

    Review the current parameters and input types below before migrating a request that depends on provider-specific features.

  3. [ 03 ]

    Treat self-hosting as a separate choice

    Moonshot says full model weights will be released by November 4, 2026. Check the published weights, license and hardware requirements before planning a self-hosted deployment.

Route

One route, standard clients.

Authenticate with a MiniRouter bearer key. Caller-provided upstream credentials are not used on this route.
Read model setup docs
Base URLhttps://api.minirouter.sh/v1
Model IDmoonshotai/kimi-k3.1-pro
Endpoints/v1/chat/completions/v1/messages/v1/responses
Parametersmax_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoning
Structured outputNot enforced; a requested format is dropped and the answer is plain text.
ModalitiesInput: text · Output: text

Configured providers. Priority follows the live catalog order.

ProviderPriorityInput / 1MOutput / 1MParametersAccess
vercel
1$2.1$10.5max_tokens, temperature, stop, tools, tool_choice, reasoning, include_reasoningMiniRouter balance

Every model, one key.

Get an API keyRead the docs