Model - Hostlinku Docs

Hostlinku provides a single centralized model catalog. Each model has a unique slug that is used in the model parameter in Chat Completions, along with context size, capabilities, and price per token.

`GET /v1/models`

Returns all models that are marked as available, not yet deprecated, and available for the Service Mode API key being used (Performance or Low-Cost), along with the price per 1M tokens in that mode.

GET https://api.hostlinku.com/v1/models
Authorization: Bearer sk-isi-api-key-kamu

Example:

curl https://api.hostlinku.com/v1/models \
  -H "Authorization: Bearer sk-isi-api-key-kamu"

Response (example truncated; the exact value for your account is retrieved from this endpoint):

{
  "object": "list",
  "service_mode": "performance",
  "data": [
    {
      "id": "claude-haiku-4-5",
      "object": "model",
      "created": 1791380432,
      "owned_by": "hostlinku",
      "context_window": 200000,
      "capabilities": ["text", "tool-calling", "function-calling", "structured-outputs", "json-mode", "streaming", "code-generation", "long-context", "multilingual"],
      "service_mode": "performance",
      "pricing": {
        "input": 0.8,
        "output": 4,
        "cache_read": 0.08,
        "cache_write": 1,
        "unit": "1M_tokens",
        "currency": "🪙"
      }
    }
  ]
}

Field	Meaning
`service_mode` (level list & item)	Service Mode API key used: `performance` or `low_cost`. The list of models varies by mode.
`id`	Model slug; the value used in the `model` field in `/v1/chat/completions`, `/v1/messages`, and `/v1/responses`.
`created`	The time the model was added to the catalog (Unix epoch, seconds).
`owned_by`	Always `"hostlinku"` — indicates that the model is managed by the Hostlinku catalog.
`context_window`	Maximum context size in tokens (input + output).
`capabilities`	List of capabilities, e.g., `vision`, `tool-calling`, `json-mode`, `reasoning`, `long-context`.
`pricing`	The pricing object for that Service Mode; `null` if the model does not yet have a price for that mode. `pricing.unit` is always `"1M_tokens"`, and `pricing.currency` contains the credit symbol (1 = $1).
`pricing.input` / `pricing.output`	Price per 1M tokens (not per token or per 1K) for prompt tokens / response tokens.
`pricing.cache_read` / `pricing.cache_write`	Price per 1M tokens for a 5-minute cache read / cache write (see Prompt Caching); `null` if the model does not support it.

Public catalog (no authentication required)

For marketing or exploratory use without an API key, a public catalog is available:

GET https://api.hostlinku.com/api/public/models

The public catalog displays only Hostlinku models along with their slugs, names, contexts, capabilities, and public prices. Supported queries: q (search by name/description), capability, family, sort (popular, newest, price-asc, price-desc, context-desc, name-asc), page, and pageSize (max 100). The data response contains items, total, page, pageSize, and totalPages.

List of Models

There are currently 45 models available. Click a model name to view its context, capabilities, price, and supported endpoints.

Model	Provider	Context	Max Output	Input /1M	Output /1M
Qwen 3.6 Plus	Alibaba	1,000,000	—	$0.294	$1.765
Qwen 3.7 Max	Alibaba	1,000,000	—	$1.765	$5.294
Baize Model	Anonymous	1,000,000	—	$1.177	$4.118
Claude Fable 5	Anthropic	1,000,000	128,000	$10	$50
Claude Fable 5.1	Anthropic	1,000,000	128,000	—	—
Claude Haiku 4.5	Anthropic	200,000	64,000	$1	$5
Claude Opus 4.6	Anthropic	1,000,000	32,000	$5	$25
Claude Opus 4.7	Anthropic	1,000,000	32,000	$5	$25
Claude Opus 4.8	Anthropic	1,000,000	128,000	$5	$25
Claude Opus 5	Anthropic	1,000,000	64,000	$5	$25
Claude Opus 5.5	Anthropic	1,000,000	128,000	$4	$20
Claude Sonnet 4.6	Anthropic	1,000,000	64,000	$3	$15
Claude Sonnet 5	Anthropic	1,000,000	64,000	$2	$10
Claude Sonnet 5.5	Anthropic	1,000,000	128,000	—	—
DeepSeek V4 Flash	DeepSeek	1,000,000	—	$0.15	$0.6
DeepSeek V4 Pro	DeepSeek	1,000,000	64,000	$0.662	$1.985
DeepSeek-V4.1-Flash	DeepSeek	1,000,000	384,000	$0.15	$0.6
DeepSeek-V4.1-Flash Hyper	DeepSeek	1,000,000	384,000	—	—
DeepSeek-V4.1-Flash Hyper Chat	DeepSeek	1,000,000	384,000	—	—
DeepSeek-V4.1-Flash Turbo	DeepSeek	1,000,000	384,000	—	—
MiniMax M3	MiniMax	1,000,000	64,000	$0.309	$1.235
Kimi K2.6	Moonshot	256,000	32,000	$0.956	$3.971
Kimi K2.7 Code	Moonshot	256,000	—	$0.956	$3.971
Kimi K3	Moonshot	1,000,000	64,000	$2.941	$14.706
Kimi K3 Hyper	Moonshot	1,000,000	64,000	—	—
Kimi K3 Hyper Chat	Moonshot	1,000,000	64,000	—	—
Kimi K3 Turbo	Moonshot	1,000,000	64,000	—	—
GPT 5.3 Codex	OpenAI	400,000	128,000	$1.75	$14
GPT-5.5	OpenAI	1,000,000	128,000	$5	$30
GPT-5.6 Luna	OpenAI	1,050,000	64,000	$0.2	$1.2
GPT-5.6 Sol	OpenAI	1,050,000	64,000	$4	$20
GPT-5.6 Terra	OpenAI	1,050,000	64,000	$2	$12
GPT-6 Astra	OpenAI	1,050,000	128,000	$10	$50
GPT-6 Luna	OpenAI	1,050,000	128,000	$0.1	$0.5
GPT-6 Sol	OpenAI	1,050,000	128,000	$2	$10
GPT-6.1 Sol	OpenAI	1,050,000	128,000	$2	$10
MiMo V2.5 Pro	Xiaomi	1,000,000	32,000	$0.441	$0.882
GLM 5	Zhipu	200,000	—	$0.588	$2.647
GLM 5.1	Zhipu	200,000	128,000	$0.882	$3.529
GLM 5.2	Zhipu	1,000,000	64,000	$1.177	$4.118
GLM 5.3	Zhipu	1,000,000	64,000	$1.177	$4.118
GLM 5.3 Hyper Chat	Zhipu	1,000,000	64,000	—	—
GLM-5.3 Flash	Zhipu	1,000,000	64,000	—	—
GLM-5.3 Flash Hyper	Zhipu	1,000,000	64,000	—	—
GLM-5.3 Flash Hyper Chat	Zhipu	1,000,000	64,000	—	—

How to Choose a Model

Determine your needs: general chat, reasoning, vision/multimodal, tool invocation, or structured JSON output. Check the capabilities field.
Pay attention to the context size: make sure context_window is large enough to accommodate your prompt and output.
Compare prices: pricing.input for prompt tokens and pricing.output for tokens Answer: Reasoning models generally have higher output costs.
Start with a small model: for development, use a lightweight model (e.g., gpt-6-luna, deepseek-v4-1-flash, qwen3-6-plus) are much cheaper; only upgrade the model if the quality is insufficient.
View the console: the Model menu at console.hostlinku.com displays a catalog with the same prices and features, plus a detail page for each model.

Routing and Failover

Hostlinku automatically manages the internal routes that serve each model. If an error occurs on the route currently being used (5xx/timeout), the request is automatically retried via another route before giving up. You continue to use the same model slug and don’t need to change anything; these internal details are not displayed in the response.

​GET /v1/models

​Public catalog (no authentication required)

​List of Models

​How to Choose a Model

​Routing and Failover

`GET /v1/models`

Public catalog (no authentication required)

List of Models

How to Choose a Model

Routing and Failover