How to Use Kimi Latest in Claude
Use Kimi Latest for large-codebase work, research synthesis, and tasks that need several tool calls. Unlike the fixed Kimi K3 name, it follows a changing model target, currently Kimi K3.
Prompts to try with Kimi Latest in Claude
Copy an example into Claude, or ask in your own words.
Use Kimi Latest to compare the three proposals pasted above. Create a table of scope, assumptions, exclusions, and unanswered questions, quoting the passage behind each finding.
Use Kimi Latest to explain the Chinese passage above in English for a product manager. Preserve product names, define technical terms, and flag phrases with more than one plausible interpretation.
Use Kimi Latest to synthesize the sources above into a 500-word decision memo. Separate established facts, disagreements, and unanswered questions, and cite each source by its supplied title.
Use Kimi Latest to analyze the incident timeline and log excerpts above. Give the most plausible explanation, the evidence against it, and three checks that would distinguish competing explanations.
Use Kimi Latest with lower reasoning effort to summarize the policy text above in 200 words. List the obligations and cite their section numbers; say when the text leaves a question unanswered.
How to set up Kimi Latest in Claude
Pick the way that suits you. Each takes about two minutes.
Use the Claude connector
- Add Spiral to Claude
Click Add Spiral in Claude above. Claude opens Add custom connector with Spiral and its URL filled in; check them and click Continue. Without the link: in Claude open Customize › Connectors › Add › Add custom connector and paste this URL:
https://api.spiral.ai/mcp
- Keep the detected sign-in and click Add
Claude finds Spiral's sign-in by itself: leave Sign in now and Use Claude's published identity selected.
- Click Connect
The connector page opens. Connect takes you to Spiral in your browser.
Use the Spiral skill
- Send Claude this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run Kimi Latest
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openrouter", "model": "~moonshotai/kimi-latest", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Claude finds Kimi Latest on Spiral, shows the price and runs it.
Kimi Latest providers and prices
Spiral runs Kimi Latest through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.058611 per run at the default inputs. $0.23465 per 1M input tokens, $14.30715 per 1M output tokens.
Settings you can ask for
Ask Claude for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 943718 | The most output tokens to generate. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Reasoning effort | none, low, high, or max | max | How much the model reasons. |
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 |
Every API field of MoonshotAI: Kimi Latest is on its tool page: open its card above.
What can you do with Kimi Latest in Claude?
Kimi Latest can turn substantial supplied text into comparisons, explanations, and research briefs. Specify the audience, evidence requirements, and answer length to give the task a clear endpoint.
Fix bugs across a large repository
Ask Kimi Latest to trace a failure through the relevant code, reproduce it, and propose a focused repair. Specify the permitted scope and require test evidence.
Plan a cross-module refactor
Have Kimi Latest map dependencies and propose a staged refactor with acceptance checks. Define the behavior that each stage must preserve.
Turn research methods into code
Give Kimi Latest a paper's methods and equations as text, then request an implementation plan, executable code, and checks against the reported results.
Compare competing proposals
Supply proposal text and ask Kimi Latest to compare scope, assumptions, exclusions, and unanswered questions. Require supporting passages for each conclusion.
Prepare a research brief
Have Kimi Latest synthesize retrieved sources into a concise brief with dated evidence, disagreements, and open questions. Define the audience and decision the brief must support.
Explain Chinese and English material
Use Kimi Latest to explain supplied Chinese material in English, or English material in Chinese. Provide a terminology list and ask it to flag ambiguous passages.
Review operational incidents
Give Kimi Latest incident timelines, log excerpts, and investigation notes as text. Ask for a causal hypothesis, conflicting evidence, and the next checks to run.
Answer questions from reference material
Supply relevant policy or product documentation as text and ask Kimi Latest to answer with section references. Require it to identify questions the material does not resolve.
More on Kimi Latest and Claude
Overview
Kimi Latest in other agents
Questions about Kimi Latest in Claude
Can Claude use Kimi Latest?
Yes. Add Spiral to Claude as an MCP server, and Claude can find and run Kimi Latest like any built-in tool.
How much does Kimi Latest cost in Claude?
Up to $0.058611 per run at the default inputs. $0.23712 per 1M input tokens, $14.30715 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude plan is separate.
Do I need an OpenRouter account to use Kimi Latest in Claude?
No. Spiral runs Kimi Latest on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Kimi Latest in Claude?
About two minutes: add Spiral to Claude, then keep the detected sign-in and click Add, then click Connect.
Does Claude wait while Kimi Latest runs?
Kimi Latest runs in the background. Claude starts the run, checks on it and gives you the text, json when it's ready.
Does Kimi Latest always mean the same model?
No. Its target can change, and the current listing points to Kimi K3. Treat saved prompts as something to recheck when the target changes.
What is Kimi Latest best suited for?
With its current Kimi K3 target, it suits sustained coding tasks, research synthesis, and knowledge work that takes several steps. Give it a concrete deliverable and a clear definition of done.
How much context can Kimi Latest handle?
The alias can reach a 1,048,576-token context window, but the actual limit depends on its target. Use that capacity for relevant evidence, and ask for a bounded answer such as a comparison table or a short decision memo.
Can Kimi Latest answer questions about today's events?
Give it current source material or have the agent retrieve it first. Kimi's training knowledge alone cannot reliably answer time-sensitive questions.
Which languages are a sensible starting point?
Chinese and English are explicitly highlighted in Moonshot's example instructions. For bilingual work, specify the output language and supply a glossary for names and technical terms.
Start using Kimi Latest in Claude
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Claude:
set up https://spiral.ai/SKILL.md


