Skip to main content

AI Coding Plans & Inference Options

Pay-as-you-go, subscription, or bring your own keys — compare every way to power Kilo with AI inference. Use the managed Kilo Gateway, plug in your own API keys, connect an existing subscription, or run models locally. No lock-in.

35
Providers & plans
500+
Models in Kilo
$0
Markup on tokens

Four ways to power Kilo

Mix and match — start managed and switch to your own keys, subscriptions, or local models anytime.

Managed by Kilo

The Kilo Gateway runs inference for you — 500+ models on a single pay-as-you-go bill, zero markup. Add Kilo Pass for bonus credits.

Bring your own keys

Plug in pay-as-you-go API keys from Anthropic, OpenAI, Google, DeepSeek, Mistral, Bedrock, and more. Keys route through the Gateway — you pay the provider at list price.

Bring your own subscriptions

Buy a coding plan like MiniMax directly in Kilo with your credits, bring one like the Z.ai Coding Plan or Kimi Code, or sign in with your ChatGPT Plus/Pro or SuperGrok / X Premium+ subscription — no API key required.

Run it locally

Connect Ollama, LM Studio, or any OpenAI-compatible endpoint to run models on your own hardware, free and offline.

Compare your options

Filter by how you want to connect. Every option works inside Kilo Code.

35 of 35 options
OptionAccessBillingPriceModels
Kilo GatewayManaged inference, zero setupRecommended
Managed by KiloPay-as-you-goPay-as-you-go500+ models from 60+ providersLearn more
Kilo PassSubscription on top of the GatewayRecommended
Managed by KiloSubscription$19 / $49 / $199 per moSame 500+ models as the GatewayLearn more
ChatGPT Plus / ProSign in with OpenAI — no API key
Subscription (sign-in)Subscription$20 / $200 per moOpenAI Codex catalog (GPT-5.5, GPT-5.3-Codex)Learn more
SuperGrok / X Premium+Sign in with xAI — no API key
Subscription (sign-in)SubscriptionSuperGrok / X Premium+ planGrok models, including Grok Build for agentic codingLearn more
AnthropicBring your own API key
Bring your own keyPay-as-you-goProvider list priceClaude Opus, Sonnet, HaikuLearn more
OpenAIBring your own API key
Bring your own keyPay-as-you-goProvider list priceGPT-5.5, GPT-4o, GPT-4o miniLearn more
Google AI StudioBring your own API key
Bring your own keyPay-as-you-goProvider list priceGemini 3 Pro, 2.5 / 2.0 FlashLearn more
xAIBring your own API key
Bring your own keyPay-as-you-goProvider list priceGrok modelsLearn more
DeepSeekBring your own API key
Bring your own keyPay-as-you-goProvider list priceDeepSeek V3, R1Learn more
Mistral AIBring your own API key
Bring your own keyPay-as-you-goProvider list priceMistral Large, CodestralLearn more
Moonshot AI (Kimi)Bring your own API key
Bring your own keyPay-as-you-goProvider list priceKimi modelsLearn more
MiniMaxBring your own API key
Bring your own keyPay-as-you-goProvider list priceMiniMax modelsLearn more
Z.aiBring your own API key
Bring your own keyPay-as-you-goProvider list priceGLM modelsLearn more
F
FireworksBring your own API key
Bring your own keyPay-as-you-goProvider list priceOpen-source models, fast hostingLearn more
Bedrock
AWS BedrockBring your own API key
Bring your own keyPay-as-you-goProvider list priceClaude, Llama & more via BedrockLearn more
N
NovitaBring your own API key
Bring your own keyPay-as-you-goProvider list priceOpen-source model hostingLearn more
I
InceptionBring your own API key
Bring your own keyPay-as-you-goProvider list priceInception modelsLearn more
X
XiaomiBring your own API key
Bring your own keyPay-as-you-goProvider list priceXiaomi modelsLearn more
O
Ollama CloudBring your own API key
Bring your own keyPay-as-you-goProvider list priceHosted open-source modelsLearn more
C
ChutesBring your own API key
Bring your own keyPay-as-you-goProvider list priceChutes-hosted modelsLearn more
C
CrofAIBring your own API key
Bring your own keyPay-as-you-goProvider list priceCrofAI-hosted modelsLearn more
I
InceptronBring your own API key
Bring your own keyPay-as-you-goProvider list priceInceptron modelsLearn more
M
MartianBring your own API key
Bring your own keyPay-as-you-goProvider list priceMartian-routed modelsLearn more
N
NeuralwattBring your own API key
Bring your own keyPay-as-you-goProvider list priceNeuralwatt modelsLearn more
O
OrcaRouterBring your own API key
Bring your own keyPay-as-you-goProvider list priceOrcaRouter modelsLearn more
S
SyntheticBring your own API key
Bring your own keyPay-as-you-goProvider list priceSynthetic-hosted modelsLearn more
MiniMax Token Plan PlusBuy in Kilo with your credits — auto-configured
Coding planSubscription$20 / mo (Kilo credits)MiniMax M3 / M2.7 (1M context), image, speech, musicLearn more
Z.ai Coding PlanCoding subscription — bring its API key
Coding planSubscriptionPlan subscriptionGLM coding modelsLearn more
Kimi CodeCoding subscription — bring its API key
Coding planSubscriptionPlan subscriptionKimi coding modelsLearn more
B
BytePlus Coding PlanCoding subscription — bring its API key
Coding planSubscriptionPlan subscriptionBytePlus coding modelsLearn more
X
Xiaomi Token Plan (Europe)Coding subscription — bring its API key
Coding planSubscriptionPlan subscriptionXiaomi token plan modelsLearn more
X
Xiaomi Token Plan (Singapore)Coding subscription — bring its API key
Coding planSubscriptionPlan subscriptionXiaomi token plan modelsLearn more
O
OpenCode GoCoding subscription — bring its API key
Coding planSubscriptionPlan subscriptionOpenCode Go modelsLearn more
OpenRouterOne key, every provider
Bring your own keyPay-as-you-goList price + small fee500+ models behind a single keyLearn more
Local (Ollama / LM Studio)Run models on your own hardware
LocalFreeFreeAny local or OpenAI-compatible modelLearn more

Prices are public list/market rates as of June 2026 and may change — check each provider for current pricing. Bring-your-own-key and coding-plan keys route through the Kilo Gateway at zero Kilo markup. Consumer subscriptions Kilo connects via sign-in (no API key) include ChatGPT Plus/Pro and SuperGrok / X Premium+, and work in the VS Code extension and CLI. Full provider list in the BYOK docs.

Inference options FAQ

Can I use my ChatGPT Plus or Pro subscription in Kilo?

Yes. Select the OpenAI ChatGPT provider and sign in with OAuth — no API key needed. Your existing ChatGPT Plus or Pro subscription covers usage and gives you access to the OpenAI Codex catalog inside Kilo.

Can I use my SuperGrok or X Premium+ subscription in Kilo?

Yes. Select the xAI provider and sign in with OAuth — no API key needed. Your SuperGrok or X Premium+ subscription gives you the latest Grok models, including Grok Build for agentic coding. Subscription sign-in works in the VS Code extension and CLI; Cloud Agents and KiloClaw use xAI via a BYOK API key instead.

Can I use a Claude Pro or Claude Max subscription the same way?

Not via sign-in. Consumer Claude subscriptions do not expose API access, so Anthropic models connect through your own Anthropic API key (BYOK). The consumer subscriptions Kilo connects via OAuth are ChatGPT Plus/Pro and SuperGrok / X Premium+.

Can I buy a coding plan directly in Kilo?

Yes. Coding plans are available in the Kilo Marketplace and are purchased with your existing Kilo balance — no Kilo Pass required. The launch plan is the MiniMax Token Plan Plus ($20/month) for MiniMax M3, which Kilo automatically configures in your BYOK settings, with more plans on the way.

Can I bring a coding-plan subscription like the Z.ai Coding Plan or Kimi Code?

Yes. Coding-focused plans such as the Z.ai Coding Plan, Kimi Code, and the BytePlus Coding Plan issue their own API key. Add that key as a BYOK provider and Kilo routes the plan’s included models through the Kilo Gateway.

Does Kilo mark up the price of tokens?

No. With Bring Your Own Key you pay the provider directly at their list price, and the Kilo Gateway adds zero markup on token usage. Kilo Pass adds bonus credits on top of Gateway usage rather than increasing prices.

Do I need an API key to get started?

No. The Kilo Gateway is the zero-setup default — sign in and use 500+ models with a single pay-as-you-go bill. API keys are optional for developers who prefer to bring their own provider billing.

Can I run models locally for free?

Yes. Kilo connects to local runtimes like Ollama and LM Studio, or any OpenAI-compatible endpoint, so you can run models offline at no token cost — ideal for private or air-gapped work.