Model List and Pricing API Reference
Use the Gate.AI model list endpoint to retrieve every available model, its core metadata, and list pricing. It is suited to model selection, comparison tools, and programmatic agent workflows. Each request returns the complete list without pagination.
| Field | Value |
|---|---|
| Base URL | https://gate.ai |
| Auth | No authentication required |
| Format | Gate.AI REST JSON |
List Models
GET
/api/v1/modelsReturns every model currently available on the platform (150+), including display names, providers, input and output modalities, context lengths, and complete list pricing.
Request Parameters
| Name | In | Type | Required | Description |
|---|---|---|---|---|
| lang | query | string | No | Response text language: en (default), zh (Simplified Chinese), or zh-tw (Traditional Chinese). This affects description and tags[].name. Values are case-sensitive; invalid values fall back to English. |
| context_length | query | string | No | Filter by context length. Supports comma-separated values, for example 128000,64000. Pass 0, an empty string, or omit the parameter for no limit. |
| input_modality | query | string | No | Filter by input modality. Supported values include text, image, and video. Multiple values can be comma-separated, for example text,image. |
| keyword | query | string | No | Fuzzy-search model names against model or base_info.name. Matching is case-insensitive. |
| max_input_price | query | number | No | Maximum input price. Pass 0 or omit the parameter for no limit. |
| max_output_price | query | number | No | Maximum output price. Pass 0 or omit the parameter for no limit. |
| min_input_price | query | number | No | Minimum input price. Defaults to 0 when omitted and uses an open lower bound, excluding models priced at 0. |
| min_output_price | query | number | No | Minimum output price. Defaults to 0 when omitted and uses an open lower bound, excluding models priced at 0. |
| output_modality | query | string | No | Filter by output modality. Supported values include text, image, and video. Multiple values can be comma-separated, for example text,image. |
| provider_name | query | string | No | Filter by the provider name visible to users. Matching ignores case and surrounding whitespace. |
Example
bash
1curl "https://gate.ai/api/v1/models?lang=zh-tw"Response Fields
| Name | Type | Description |
|---|---|---|
| code | integer | Business status code. Success is always 200. |
| message | string | Status description. Success is success. |
| data.data | array | Model array. Each element is a model object. |
| timestamp | integer | Server timestamp in milliseconds. |
| data.data[].model | string | Model ID to pass as the model parameter to generation endpoints. |
| data.data[].name | string | Display name. |
| data.data[].provider_name | string | Model provider name, such as OpenAI, Anthropic, or ByteDance. |
| data.data[].type | string | Supported endpoint forms, separated by |
| data.data[].input_price | number | null |
| data.data[].output_price | number | null |
| data.data[].cache_read_price | number | null |
| data.data[].cache_write_price | number | null |
| data.data[].context_length | integer | Context length in tokens. Models for which this does not apply return 0. |
| data.data[].description | string | Model description localized by the lang parameter. |
| data.data[].tags | array | Capability tags containing key (stable English identifier), name (localized by lang), and category. |
| data.data[].input_modalities | array | Input modalities: text, image, video, or audio. This is platform-maintained reference data and may lag individual upstream model updates. |
| data.data[].output_modalities | array | Output modalities, using the same value set as input_modalities. |
| data.data[].pricing_tiers | object | Present only for tier-priced models. input, output, cache_read, and cache_write contain tier arrays with an upper bound and the corresponding price. |
| data.data[].video_capabilities | object | Present only for video models: durations (seconds), resolutions, and aspect_ratios. |
| data.data[].video_pricing | object | Pricing details present only for video models. |
| video_pricing.currency | string | Currency. Currently USD. |
| video_pricing.billing_mode | string | per_second_resolution (output seconds × resolution tier) or per_film (per generated video). |
| video_pricing.unit | string | Billing unit: second or film. |
| video_pricing.tiers | array | Pricing tiers. Per-second entries contain resolution + price; per-film entries contain resolution + duration + price. Prices are strings. |
| data.data[].image_pricing | object | Pricing details present only for image models. |
| image_pricing.currency | string | Currency. Currently USD. |
| image_pricing.billing_mode | string | per_image or token. |
| image_pricing.unit | string | Billing unit: image or 1M. |
| image_pricing.per_image | string | Price per image for per-image billing. |
| image_pricing.text_input / image_input | object | Present for token billing. Each object contains input, output, and cache_read prices in USD per million tokens as strings. |
| data.data[].tts_stt_pricing | object | Pricing details present only for speech models. |
| tts_stt_pricing.billing_mode | string | token or audio_duration. |
| tts_stt_pricing.input_tokens / output_tokens / cache_read_tokens / cache_write_tokens | string | Present for token billing, in USD per million tokens. |
| tts_stt_pricing.input_per_minute | string | Present for duration billing. Input-audio price per minute in USD. |
| data.data[].model_publish_time | string (ISO 8601) | Model publication time. |
Response Example
json
1{2 "code": 200,3 "message": "success",4 "data": {5 "data": [6 {7 "id": 224,8 "model": "qwen3-vl-plus",9 "name": "Qwen3 VL Plus",10 "icon": "https://gimg2.staticimgs.com/image/qwen_20260402_161657_e06ffb27e1219c85184d939bce7cd1bd.webp",11 "provider_type": "qwen",12 "provider_name": "Qwen",13 "type": "completions",14 "input_price": 0.144,15 "output_price": 1.434,16 "cache_read_price": 0.029,17 "cache_write_price": 0.539,18 "price_type": 0,19 "context_length": 262144,20 "description": "The flagship Plus vision-language model in the Qwen3-VL series. It supports joint image, video, and text input, and excels at document and chart parsing, video understanding, spatial localization, and intelligent agent tool calling. It is suitable for multimodal question answering and visual automation scenarios.",21 "overview": "",22 "tags": [23 {24 "id": 3,25 "key": "vision",26 "name": "Vision",27 "category": "general"28 },29 {30 "id": 28,31 "key": "ui-understanding",32 "name": "UI Understanding",33 "category": "general"34 },35 {36 "id": 30,37 "key": "document-ocr",38 "name": "Document OCR",39 "category": "general"40 },41 {42 "id": 31,43 "key": "multimodal-understanding",44 "name": "Multimodal Understanding",45 "category": "general"46 }47 ],48 "input_modalities": [49 "text",50 "image",51 "video"52 ],53 "output_modalities": [54 "text"55 ],56 "playground_enabled": true,57 "model_publish_time": "2025-09-23T00:00:00Z",58 "create_time": "2026-06-05T14:18:56Z",59 "update_time": "2026-07-06T01:57:19Z",60 "pricing_tiers": {61 "input": [62 {63 "upper": 32000,64 "input_price": "0.144"65 },66 {67 "upper": 128000,68 "input_price": "0.216"69 },70 {71 "upper": null,72 "input_price": "0.431"73 }74 ],75 "output": [76 {77 "upper": 32000,78 "output_price": "1.434"79 },80 {81 "upper": 128000,82 "output_price": "2.151"83 },84 {85 "upper": null,86 "output_price": "4.301"87 }88 ],89 "cache_read": [90 {91 "upper": 32000,92 "cache_read_price": "0.029"93 },94 {95 "upper": 128000,96 "cache_read_price": "0.044"97 },98 {99 "upper": null,100 "cache_read_price": "0.087"101 }102 ],103 "cache_write": [104 {105 "upper": null,106 "cache_write_price": "0.539"107 }108 ]109 }110 }111 ]112 },113 "timestamp": "1784201129"114}Response
| Status | Meaning | Description | Schema |
|---|---|---|---|
| 200 | OK | Success. Returns the complete model list. | ModelsResponse |
| 500 | Internal Server Error | Internal server error. | ErrorResponse |
Compatibility notice: Response fields not listed on this page are internal platform fields and may change at any time; do not depend on them. Fields documented here may be added to in a backward-compatible way but will not receive breaking changes.