Models — Hypery Docs

Overview

Access 2000+ models from multiple providers through one unified API:

Vercel AI Gateway - Chat models and the default provider (Claude, GPT, Gemini, Qwen, etc.)
OpenRouter - Chat models (GPT, Claude, Gemini, Llama, etc.)
Replicate - Image/video generation models
More coming soon

Authentication: Optional (public browsing allowed)
Required Scopes (OAuth): ai:models (recommended)

The Models API uses OpenAI-compatible format, so it works with the OpenAI SDK.

List All Models

Get all available models.

Endpoint: GET /api/v1/models

import OpenAI from 'openai'
 
const client = new OpenAI({
  apiKey: process.env.GATEWAY_API_KEY,
  baseURL: 'https://hypery.ai/api/v1'
})
 
const models = await client.models.list()
console.log(models.data)  // Array of models

const response = await fetch('https://hypery.ai/api/v1/models', {
  headers: { 'Authorization': 'Bearer YOUR_API_KEY' }
})
 
const data = await response.json()
console.log(data.data)  // Array of models

from openai import OpenAI
 
client = OpenAI(
    api_key='your-api-key',
    base_url='https://hypery.ai/api/v1'
)
 
models = client.models.list()
for model in models.data:
    print(model.id)

import requests
 
response = requests.get(
    'https://hypery.ai/api/v1/models',
    headers={'Authorization': 'Bearer YOUR_API_KEY'}
)
 
data = response.json()
for model in data['data']:
    print(model['id'])

curl https://hypery.ai/api/v1/models \
  -H "Authorization: Bearer YOUR_API_KEY"

Response:

{
  "object": "list",
  "data": [
    {
      "id": "anthropic/claude-3.5-sonnet",
      "object": "model",
      "owned_by": "anthropic",
      "name": "Claude Sonnet 4.5",
      "description": "Anthropic's most advanced Sonnet model…",
      "provider": "openrouter",
      "capabilities": {
        "function_calling": true,
        "vision": true
      },
      "context_length": 200000,
      "max_tokens": 4096
    }
  ],
  "has_more": true,
  "meta": { "total": 214, "offset": 0, "limit": 20, "hasMore": true, "returned": 20 }
}

Both response formats include pagination: the default (OpenAI) format adds has_more (OpenAI-style) plus a meta block, and format=custom returns the same meta. Use meta.total / has_more to decide whether to fetch the next offset — you don't need to switch formats to paginate.

Model Fields

Field	Type	Description
`id`	string	Model identifier (e.g., `anthropic/claude-3.5-sonnet`)
`name`	string	Human-readable display name (gateway extension — the OpenAI spec omits this)
`description`	string	Short model description (gateway extension)
`owned_by`	string	Model creator/organization
`provider`	string	Which provider hosts it (`vercel`, `openrouter`, `replicate`)
`capabilities`	object	Features (function calling, vision, etc.)
`context_length`	number	Max context window in tokens
`max_tokens`	number	Max output tokens

name and description are gateway extensions. The base OpenAI /v1/models object is minimal (id, object, created, owned_by) — clients that want a display label should read name (falling back to id).

The unfiltered list is sorted by popularity, which is dominated by image/video models. If you want a specific kind of model — most commonly chat — filter with category (below) instead of scanning the full catalog.

Filter by Category (chat / image / video …)

The fastest way to get the models you actually want. category=chat returns the LLM chat models (from Vercel AI Gateway and OpenRouter) with no image/video noise.

// All chat models (the usual case) — bump the limit to get the whole set
const response = await fetch(
  'https://hypery.ai/api/v1/models?category=chat&limit=1000'
)
const { data } = await response.json()

response = requests.get(
    'https://hypery.ai/api/v1/models',
    params={'category': 'chat', 'limit': 1000},
)
models = response.json()['data']

curl "https://hypery.ai/api/v1/models?category=chat&limit=1000"

Categories: chat · image · video · audio · embedding

Filter by Capabilities

Find models with specific features.

// Get models with function calling
const response = await fetch(
  'https://hypery.ai/api/v1/models?capabilities=function_calling',
  { headers: { 'Authorization': 'Bearer YOUR_API_KEY' } }
)
 
const models = await response.json()

# Get models with vision
response = requests.get(
    'https://hypery.ai/api/v1/models',
    params={'capabilities': 'vision'},
    headers={'Authorization': 'Bearer YOUR_API_KEY'}
)
 
models = response.json()

# Get models with multiple capabilities
curl "https://hypery.ai/api/v1/models?capabilities=function_calling,vision" \
  -H "Authorization: Bearer YOUR_API_KEY"

Available capabilities:

function_calling - Can use tools
vision - Can analyze images
reasoning - Extended reasoning (o1 models)

Search Models

Find models by name or description.

const response = await fetch(
  'https://hypery.ai/api/v1/models?search=claude',
  { headers: { 'Authorization': 'Bearer YOUR_API_KEY' } }
)
 
const models = await response.json()

response = requests.get(
    'https://hypery.ai/api/v1/models',
    params={'search': 'claude'},
    headers={'Authorization': 'Bearer YOUR_API_KEY'}
)
 
models = response.json()

curl "https://hypery.ai/api/v1/models?search=claude" \
  -H "Authorization: Bearer YOUR_API_KEY"

Searches in: model ID, name, and description.

Filter by Provider

Get models from a specific provider.

// Vercel AI Gateway — the DEFAULT chat provider (Claude, GPT, Gemini, Qwen, …)
const response = await fetch(
  'https://hypery.ai/api/v1/models?provider=vercel&limit=1000'
)
 
// OpenRouter chat models
const response = await fetch(
  'https://hypery.ai/api/v1/models?provider=openrouter&limit=1000'
)
 
// Replicate models (images/video)
const response = await fetch(
  'https://hypery.ai/api/v1/models?provider=replicate'
)

# Vercel AI Gateway (default chat provider)
curl "https://hypery.ai/api/v1/models?provider=vercel&limit=1000"
 
# OpenRouter chat models
curl "https://hypery.ai/api/v1/models?provider=openrouter&limit=1000"
 
# Replicate models
curl "https://hypery.ai/api/v1/models?provider=replicate"

Filter by Context Length

Find models with large context windows.

// Models with at least 100K tokens
const response = await fetch(
  'https://hypery.ai/api/v1/models?minContextLength=100000'
)

curl "https://hypery.ai/api/v1/models?minContextLength=100000"

Filter by pricing tier

By default, the public catalog exposes pricing tiers (free, budget, standard, premium) — not per-token unit costs. Use pricingTier to filter:

// Budget-tier models
const response = await fetch(
  'https://hypery.ai/api/v1/models?format=custom&pricingTier=budget'
)

curl "https://hypery.ai/api/v1/models?format=custom&pricingTier=budget"

Numeric minCost / maxCost filters (per 1M tokens) are only available when your OAuth app is configured for transparent pricing presentation. The default is tiered.

Pagination

Handle large result sets. Read has_more (or meta.total) from the response to decide whether to fetch the next offset. A single page can return up to limit=1000, so most filtered slices (e.g. ?category=chat) fit in one request.

// First page (50 results)
const page1 = await fetch(
  'https://hypery.ai/api/v1/models?limit=50&offset=0'
)
 
// Second page
const page2 = await fetch(
  'https://hypery.ai/api/v1/models?limit=50&offset=50'
)

# First page
curl "https://hypery.ai/api/v1/models?limit=50&offset=0"
 
# Second page
curl "https://hypery.ai/api/v1/models?limit=50&offset=50"

Query Parameters

Parameter	Type	Description	Example
`category`	string	Filter by model kind — `chat`, `image`, `video`, `audio`, `embedding`	`chat`
`provider`	string	Filter by provider	`vercel`, `openrouter`, or `replicate`
`capabilities`	string	Filter by features (comma-separated)	`function_calling,vision`
`inputModality`	string	Filter by accepted input(s) (comma-separated)	`text,image`
`outputModality`	string	Filter by produced output(s) (comma-separated)	`text`
`search`	string	Search query (id, name, description)	`claude`
`minContextLength`	number	Minimum context window	`100000`
`pricingTier`	string	Filter by tier (`free`, `budget`, `standard`, `premium`)	`budget`
`maxCost`	number	Max cost per 1M tokens (transparent only)	`1.0`
`minCost`	number	Min cost per 1M tokens (transparent only)	`0.5`
`limit`	number	Results per page — default `20`, max `1000` (values above are clamped)	`1000`
`offset`	number	Skip results (for pagination)	`100`

limit is capped at 1000 per request, and the unfiltered catalog is sorted by popularity — so the first page is dominated by high-traffic image/video models. Always filter (e.g. ?category=chat) rather than paging the whole catalog to find a model kind. To enumerate a large slice, paginate with offset.

Model ID Format

Models use the format: organization/model-name

Examples:

anthropic/claude-3.5-sonnet
openai/gpt-4o
google/gemini-pro
meta-llama/llama-3-70b
black-forest-labs/flux-schnell

The first part is the creator/organization, NOT the provider. Provider is a separate field.

Popular Models

Chat Models

Model ID	Context	Best For
`anthropic/claude-3.5-sonnet`	200K	Coding, analysis
`openai/gpt-4o`	128K	General purpose
`openai/gpt-4o-mini`	128K	Fast, cheap
`google/gemini-pro`	32K	Speed

Image Models

Model ID	Speed	Best For
`black-forest-labs/flux-schnell`	~15s	High quality
`stability-ai/stable-diffusion-3`	~20s	General purpose

View all models in dashboard →

Capabilities Reference

Capability	What It Means
`function_calling`	Model can use tools/functions
`vision`	Model can analyze images
`reasoning`	Extended thinking (o1 models)

Filter example:

# Get models that can see images AND call functions
GET /api/v1/models?capabilities=vision,function_calling

Next Steps

Chat API

Use models for chat completions

Predictions API

Use models for async image generation

Providers

Learn about providers