ai-model-prices

Pricing data for AI models, typed and kept up to date daily from models.dev.

226 providers · 8538 models · Updated 2026-10-31

Installation

npm install ai-model-prices

Usage

Main package

import { getPrice, findModel, findProvider, getModels } from 'ai-model-prices';

// By provider ID + model ID
getPrice('openai/gpt-4.1-mini', { noCacheInput: 1000, noCacheOutput: 500 });

// Two call signatures
getPrice('openai', 'gpt-4.1-mini', { noCacheInput: 1000, noCacheOutput: 500 });
getPrice('amazon-bedrock', 'claude-3-5-sonnet-20241022-v2:0', { noCacheInput: 500, noCacheOutput: 200 });

// Get a model
const model = getModel('openai', 'gpt-4.1-mini');

// Get a provider
const provider = getProvider('amazon-bedrock');

Provider adapters

Each provider has a lightweight adapter. Types are scoped to that provider's models only:

import { getPrice, getModel, getProvider, getModels } from 'ai-model-prices/openai';
import { getPrice as bedrockPrice } from 'ai-model-prices/amazon-bedrock';

// provider is pre-bound
const price = getPrice('gpt-4.1-mini', { noCacheInput: 1000, noCacheOutput: 500 });
const model = getModel('gpt-4.1-mini'); // typed to OpenAI models only

Async (lazy-loaded)

If bundle size matters, provider data is loaded on demand:

import { getPrice, getModels, filterModels } from 'ai-model-prices/async';

const models = await filterModels({ maxInputCost: 1, tool_call: true });

API

getPrice(model, tokens)

Returns a PriceBreakdown, or null if the model has no pricing data.

// Three call signatures:
getPrice(model: Model, tokens: TokenUsage): PriceBreakdown | null
getPrice('providerId/modelId', tokens: TokenUsage): PriceBreakdown | null
getPrice('providerId', 'modelId', tokens: TokenUsage): PriceBreakdown | null

interface TokenUsage {
  noCacheInput?: number;   // regular input tokens
  noCacheOutput?: number;  // output tokens
  cacheInput?: number;     // cache read tokens
  cacheWrite?: number;     // cache write tokens
  reasoning?: number;      // reasoning tokens (o1, etc.)
  inputAudio?: number;
  outputAudio?: number;
}

interface PriceBreakdown {
  noCacheInput: number;
  noCacheOutput: number;
  cacheInput: number;
  cacheWrite: number;
  reasoning: number;
  audio: number;
  total: number;           // USD
}

getModel(providerId, modelId)

Both arguments are typed to known IDs.

getProvider(id)

Look up a provider by ID.

getModels() / getProviders() / getModelsByProvider(providerId)

Return everything from the bundled data.

filterModels(filter)

filterModels({
  provider: 'openai',           // one or more provider IDs
  tool_call: true,
  reasoning: false,
  maxInputCost: 5,              // USD per 1M tokens
  maxOutputCost: 20,
  minContextWindow: 128_000,
  inputModalities: ['image'],
})

Supported Providers

Provider	ID
OpenReason	openreason
DaoXE	daoxe
Chutes	chutes
Anthropic	anthropic
Vercel AI Gateway	vercel
Melious	melious
STACKIT	stackit
klokintegration.se	klokintegration
Munito AI (EU)	munito-eu
Perplexity	perplexity
Bothub	bothub
InferX	inferx
Novita AI	novita-ai
NanoGPT	nano-gpt
SAP AI Core	sap-ai-core
Jiekou.AI	jiekou
Subconscious	subconscious
Vertex	google-vertex
SCNet Token Plan	scnet-token-plan
Berget.AI	berget
Inco	inco
Google	google
Thinking Machines	thinkingmachines
Echo	echo
Amazon Bedrock	amazon-bedrock
HPC-AI	hpc-ai
LMStudio	lmstudio
Ambient	ambient
Ofox	ofox
Infomaniak	infomaniak
Vivgrid	vivgrid
SiliconFlow (China)	siliconflow-cn
Clarifai	clarifai
ai&	aiand
Tencent Token Plan	tencent-token-plan
Sarvam AI	sarvam
Hopper	hopper
Z.AI Coding Plan	zai-coding-plan
Ollama Cloud	ollama-cloud
KUAE Cloud Coding Plan	kuae-cloud-coding-plan
QiHang	qihang-ai
AMD	amd
Alibaba Coding Plan	alibaba-coding-plan
LongCat	longcat
Mixlayer	mixlayer
Eden AI	edenai
Crusoe	crusoe
Hetzner	hetzner
IO.NET	io-net
Bee by HEOSSI	bee
Reflection	reflection
ClinePass	cline-pass
CoralBricks	coralbricks
Tencent Coding Plan (China)	tencent-coding-plan
Fireworks AI	fireworks-ai
Databricks	databricks
OpenRouter	openrouter
OpenCode Go	opencode-go
Requesty	requesty
Alibaba Coding Plan (China)	alibaba-coding-plan-cn
OpenAI	openai
DevPass (LLM Gateway)	llmgateway
Kenari	kenari
Xiaomi	xiaomi
Meta	meta
Neuralwatt	neuralwatt
Cloudflare AI Gateway	cloudflare-ai-gateway
Pareto Inference	pareto
Lilac	lilac
Morph	morph
Aixy	aixy
MiniMax (minimax.io)	minimax
MiniMax Token Plan (minimax.cn)	minimax-cn-coding-plan
Pendra	pendra
Deep Infra	deepinfra
TokenGo	tokengo
Kimi For Coding (kimi.com)	kimi-code-plan-cn
IteraCompute	iteracompute
Synthetic	synthetic
Prime Inference	prime
Wallaby	wallaby
Upstage	upstage
SambaNova	sambanova
Tencent TokenHub	tencent-tokenhub
GitLab Duo	gitlab
Umans AI	umans-ai
Regolo AI	regolo-ai
Cerebras	cerebras
EBCloud	ebcloud
Modal	modal
Impossibl	impossibl
ainetcafe	ainetcafe
Cortecs	cortecs
RunInfra	runinfra
Venice AI	venice
Poolside	poolside
CrossModel	crossmodel
Charm Hyper	hyper
Alibaba Token Plan	alibaba-token-plan
SaladCloud AI Gateway	salad-cloud
Infer by Flow7	infer
Kimi For Coding (kimi.ai)	kimi-code-plan-global
Mistral	mistral
Claudinio	claudinio
Cloudflare Workers AI	cloudflare-workers-ai
Vispark	vispark
Poe	poe
xAI	xai
Volcengine Ark	volcengine
Jalapeno Cloud	jalapeno
QVAC	qvac
Auriko	auriko
MiniMax Token Plan (minimax.io)	minimax-coding-plan
TokenRouter	tokenrouter
Kosmik Compute	kosmik
Qiniu	qiniu-ai
NEAR AI Cloud	nearai
NaN	nan
AI-ROUTER	ai-router
OCI Generative AI	oci
Moonshot AI	moonshotai
302.AI	302ai
Arcee	arcee
StepFun (Global)	stepfun-ai
above.dev	above
FastRouter	fastrouter
Umans AI Coding Plan	umans-ai-coding-plan
Z.AI	zai
abliteration.ai	abliteration-ai
Kilo Gateway	kilo
Volcengine Ark Coding Plan	volcengine-coding-plan
LLMTR	llmtr
Lynkr	lynkr
CloudFerro Sherlock	cloudferro-sherlock
Nebius Token Factory	nebius
Meganova	meganova
NeoSmith	neosmith
Xiaomi Token Plan (Europe)	xiaomi-token-plan-ams
Nova	nova
SenseNova (China)	sensenova
routing.run	routing-run
Azure	azure
LowRouter	lowrouter
Modular	modular
Standard Compute	standardcompute
Xiaomi Token Plan (China)	xiaomi-token-plan-cn
Privatemode AI	privatemode-ai
Zenifra	zenifra
Helicone	helicone
AI/ML API	aimlapi
DInference	dinference
Pioneer	pioneer
Xiaomi Token Plan (Singapore)	xiaomi-token-plan-sgp
Merge Gateway	merge-gateway
ModelScope	modelscope
Atomic Chat	atomic-chat
GMI Cloud	gmicloud
Wally	wally
watsonx.ai	watsonx
Baseten	baseten
evroc	evroc
Bailing	bailing
LucidQuery	lucidquery
Vultr	vultr
StepFun (China)	stepfun
Munito AI	munito
Zhipu AI Coding Plan	zhipuai-coding-plan
Hugging Face	huggingface
Vancine	vancine
Zhipu AI	zhipuai
Perplexity Agent	perplexity-agent
AKI.IO	aki-io
Agnes AI	agnes
DeepSeek	deepseek
FreeModel	freemodel
Snowflake Cortex	snowflake-cortex
Inception	inception
TensorX	tensorx
AIHubMix	aihubmix
Opper	opper
CommandCode	commandcode
Alibaba	alibaba
Azure Cognitive Services	azure-cognitive-services
OVHcloud AI Endpoints	ovhcloud
submodel	submodel
SCX.ai	scx-ai
Groq	groq
LLM Tech	llmtech
Together AI	togetherai
Inference	inference
Xpersona	xpersona
Vertex (Anthropic)	google-vertex-anthropic
SiliconFlow	siliconflow
engy	engy
Alibaba Token Plan (China)	alibaba-token-plan-cn
EmpirioLabs AI	empiriolabs
CoreWeave	wandb
Zeldoc	zeldoc
Moark	moark
Neon	neon
Moonshot AI (China)	moonshotai-cn
Scaleway	scaleway
D.Run (China)	drun
GreenPT	greenpt
GitHub Copilot	github-copilot
Friendli	friendli
Alibaba (China)	alibaba-cn
FrogBot	frogbot
Wafer	wafer.ai
Nvidia	nvidia
OpenCode	opencode
AI21 Labs	ai21
ZenMux	zenmux
UnoRouter	unorouter
Modelis	modelis
OrcaRouter	orcarouter
Tinfoil	tinfoil
DigitalOcean	digitalocean
Boundless	boundless
Abacus	abacus
LLM Gateway	llmgateway-providers
MiniMax (minimax.cn)	minimax-cn
Cohere	cohere
Tempr Gateway	tempr
Sakana AI	sakana
Inceptron	inceptron

License

MIT