How to Use Kimi K3.1 Pro for Free ?

If you want to use Kimi K3.1 Pro for free, there’s one verified $0 route as of October 24, 2026: the Kimi web and mobile chat with a free account, where K3.1 Pro is already the default model. The Kimi K3.1 Pro free API doesn’t exist. The API needs a $1 minimum top-up, Kimi Code needs a paid membership, and no third-party router lists a free K3.1 Pro variant.

Below: every route we checked, what each costs, and how to stretch a $1 top-up. Every price and limit below comes from Moonshot AI’s own pages, read on October 24, 2026: the membership pricing page, the API pricing page, the recharge and rate limits page and the K3.1 Pro announcement.

New to the model? Start with what Kimi K3.1 Pro is, then come back here.

Is Kimi K3.1 Pro free?

Partly. Kimi K3.1 Pro is free to chat with on the Kimi web app and mobile apps under the free Adagio plan, where it’s the default model. It isn’t free anywhere else: the API, Kimi Code, OpenRouter and other gateways all charge for it, and the open weights aren’t out yet.

Here’s what we saw on October 24, 2026:

• Free: Kimi web and app chat on the Adagio plan ($0).
• Paid: the Kimi API ($1 minimum top-up), Kimi Code (paid membership), OpenRouter ($2 / $10 per 1M tokens).
• Not yet: open weights, which Moonshot says it’ll release “by Nov 4.”

If you read our earlier guide on using Kimi K3 for free, note the difference. K3.1 Pro is a separate, newer model, released October 23, 2026, and its access rules on the API are stricter.

Every Kimi K3.1 Pro route compared

This table lists every route we checked, the cost, whether money has to go in first, and whether K3.1 Pro is actually available there.

Only the first two rows are confirmed $0 routes to K3.1 Pro today.

How to use Kimi K3.1 Pro for free in the web app

You can start in under a minute:

1. Open Kimi in your browser, or install the Kimi app on iOS, Android or HarmonyOS.
2. Check the model picker in the composer. On October 24 it showed “K3.1 Pro” as the default, with thinking effort set to “High.” The other options were K2.8 Preview and Instant.
3. Sign in with a free account. You can see the composer logged out, but sending a prompt asks you to sign in.
4. Type your prompt and send it.

No API key, no plan choice.

Kimi K3.1 Pro free tier limits

Moonshot doesn’t publish how many K3.1 Pro messages the free plan allows. The membership page lists the Adagio plan’s other limits, but no K3.1 Pro message count. [VERIFY] Treat any number you see quoted elsewhere as unconfirmed.

What the free Adagio plan does list, per the membership page on October 24:

Moonshot lists 1M-token chat context only for Max and Ultra, so free chat doesn’t get the full 1,048,576-token window. K3.1 Pro Highspeed (“2-3x the speed of K3.1 Pro”) starts at the Pro plan.

Is there a Kimi K3.1 Pro free API?

No. Moonshot’s quickstart states it directly: “Kimi K3.1 Pro is a flagship model: it is unlocked after a successful top-up (minimum $1).” You can create an account and an API key without paying, but calls to kimi-k3.1-pro won’t work until that first top-up lands.

A few details matter here:

• Rate limits follow your top-up. Your cumulative recharge sets your tier. Per the limits page on October 24, a $1 account sits in Tier0: 1 concurrent request, 3 requests per minute, 500,000 tokens per minute and 1,500,000 tokens per day. Tier1 starts at $10 cumulative and lifts you to 100 RPM with no daily token cap.
• The $5 voucher. When your cumulative recharge reaches $5, Moonshot adds a $5 voucher. Vouchers don’t count toward your cumulative recharge.
• Vouchers and K3.1 Pro. Moonshot’s official OpenCode guide says vouchers from new-user verification “cannot be used for Kimi K3.” It doesn’t say whether the rule extends to K3.1 Pro. Don’t plan around a voucher until you’ve confirmed it in your console. [VERIFY]

Gateways don’t change that: OpenRouter charges the same $2 / $10 with no :free variant.

For the full paid setup, including Codex and Claude Code config, read how to use Kimi K3.1 Pro.

Kimi K3.1 Pro open weights: free to download, not free to run

Moonshot’s announcement says “the full model weights will be released by Nov 4.” Until then there’s nothing to download. The license hasn’t been stated either, so whether self-hosting is free of license fees is still open. [VERIFY]

The bigger cost is hardware. K3.1 Pro shares the same 2.8T-parameter MoE design as K3 (about 104B active), and the K3 weights were 594 GB in MXFP4 when they came out in July. You’ll need a multi-GPU server to hold that, which is not a free resource for most people.

Our guide to running Kimi K3 locally shows what that took for K3.

Cheapest way to use Kimi K3.1 Pro: the $1 top-up

The cheapest paid route is a $1 top-up on the Kimi API. At $2 per 1M input tokens and $10 per 1M output tokens, $1 buys about 500K input tokens or about 100K output tokens. Real requests mix both, so your mileage sits between those two numbers.

Here’s the official price list from the pricing page, per 1M tokens:

K3.1 Pro is a third cheaper than K3 on both input and output. Three habits stretch your dollar further:

1. Reuse your prefix. Caching is automatic for repeated prefixes, with a default TTL of 5 minutes. Cached input costs $0.20 per 1M, a tenth of the normal rate. Keep your system prompt and shared context identical at the start of every request.
2. Set reasoning_effort to "low". K3.1 Pro always thinks, and the default is "max". Moonshot’s reasoning effort guide describes the setting as a way to balance reasoning depth, latency and token usage. Use "low" for simple tasks.
3. Stay on the standard model. Highspeed doubles both input and output prices.

A minimal low-effort call, adapted from Moonshot’s quickstart:

curl https://api.moonshot.ai/v1/chat/completions \
  --header "Authorization: Bearer $MOONSHOT_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{"model": "kimi-k3.1-pro","reasoning_effort": "low","messages": [{"role": "user", "content": "Introduce Kimi K3.1 Pro in one sentence."}]}'

Get your key at the API keys console after the top-up.

Don’t burn your $1 on broken requests: test with Apidog CLI

A $1 balance doesn’t survive many retries from a malformed body. Let your coding agent build one test case, validate it, and run it once. Apidog CLI gives Claude Code, Codex, Cursor or OpenCode the commands to do that, and Apidog’s Free plan covers up to 4 users.

Install the CLI and the agent skill:

npm install -g apidog-cli@latest
apidog login --with-token <YOUR_TOKEN>
apidog skill install

Then give your agent one prompt:

In Apidog project <projectId>, create an environment for https://api.moonshot.ai/v1.
Build a test case for POST /chat/completions with model "kimi-k3.1-pro",
reasoning_effort "low" and a small max_tokens. Send the key from MOONSHOT_API_KEY
as a Bearer token. Assert status 200 and a non-empty choices[0].message.content.
Validate the payload, then run the case once and summarize the result.

The agent reads the schema, validates its payload, creates the case and runs it a single time:

apidog environment create kimi-prod --project <projectId> --base-url https://api.moonshot.ai/v1
apidog cli-schema get test-case-create
apidog cli-schema validate test-case-create --file case.json
apidog test-case create --project <projectId> --file case.json
apidog run --test-case <caseId> --project <projectId> -e <envId> -r cli,json

cli-schema validate catches a bad payload before anything hits Moonshot’s endpoint, so a typo costs you nothing. Test cases need a valid category ID, and apidog test-case category --project <projectId> lists them. Once the case passes, the same apidog run with -r junit drops into CI; our Apidog CLI in Claude Code and CI/CD pipeline guides walk through both.

Which route should you pick?

Pick by what you’re building:

• Chatting and testing ideas: the free web or app chat.
• Building an app or agent: the API with a $1 top-up and reasoning_effort at "low".
• Coding all day in a terminal: a paid Kimi Code membership (see our Kimi Code CLI guide).
• Your own hardware: wait for the weights and check the license.

FAQ

Can I use Kimi K3.1 Pro for free without an account?

No. On October 24, 2026, the Kimi web chat showed K3.1 Pro in the model picker before login, but sending a prompt required signing in with a free account.

How many free Kimi K3.1 Pro messages do I get per day?

Moonshot doesn’t publish that number. The membership page lists the free plan’s task, project and storage limits, but no K3.1 Pro message quota. [VERIFY]

Does the Kimi API have a free tier for K3.1 Pro?

No. K3.1 Pro unlocks after a successful top-up of at least $1. At $2 / $10 per 1M tokens, that buys about 500K input or 100K output tokens.

Is Kimi K3.1 Pro free on OpenRouter?

No. OpenRouter lists moonshotai/kimi-k3.1-pro at $2 input and $10 output per 1M tokens, with a 1,048,576-token context and no free variant, as of October 24, 2026.

Does Kimi Code include K3.1 Pro on the free plan?

No. Kimi Code is a paid membership benefit. The k3.1-pro model needs Plus or above, and the full 1M context in Kimi Code needs Pro or above.

When can I self-host Kimi K3.1 Pro?

Moonshot says the full weights will be released by November 4, 2026. The license isn’t stated yet [VERIFY], and you’ll still pay for hardware: K3’s weights were 594 GB.