API Address Information

Main API Addresses

1. OpenAI SDK Compatible Interface Recommended

Item	Details
Base URL	https://api.agentsflare.com/v1
Full Address Example	https://api.agentsflare.com/v1/chat/completions
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	OpenAI (including gpt-6.1-terra), Gemini (SDK method), Grok, Deepseek (including deepseek-flash), Kimi (including kimi-k3.1-pro-highspeed and kimi-k3.1-pro), GLM (including glm-5.3 and glm-5.3-flash), etc.
Service Status	Stable

Model List Interface

Endpoint	Description	Use Case
https://api.agentsflare.com/v1/models	List Models	Retrieve all available models for the current account

Responses Interface

Endpoint	Description	Use Case
https://api.agentsflare.com/v1/responses	Responses API	Calling Codex series models (gpt-5.3-codex, gpt-5.2-codex, etc.) and Pro models; supports function calling, code generation, and other advanced features

Embeddings Interface

Endpoint	Description	Use Case
https://api.agentsflare.com/v1/embeddings	Embeddings API	Text vectorization with text-embedding-v4; supports dimensions and encoding_format parameters. See text-embedding-v4 API Examples

API Key Usage Statistics Interface

Use this endpoint to query usage associated with the API Key supplied in the Authorization header for a specified time range.

Item	Details
Endpoint	https://api.agentsflare.com/v1/usage-stats
Method	GET
Authentication	Authorization: Bearer $TOKEN
from	Start time in ISO 8601 format, including the time-zone offset
to	End time in ISO 8601 format, including the time-zone offset

curl --location --get 'https://api.agentsflare.com/v1/usage-stats' \
  --data-urlencode 'from=2026-08-26T00:00:00+08:00' \
  --data-urlencode 'to=2026-08-26T23:00:00+08:00' \
  --header "Authorization: Bearer $TOKEN"

The API Key is identified from the Bearer Token; no separate Key ID parameter is required. --data-urlencode ensures that the + sign in the time-zone offset is encoded correctly.

2. Anthropic API Interface Recommended

Item	Details
Base URL	https://api.agentsflare.com/anthropic/v1/messages
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	Claude (including claude-haiku-5-5)
Service Status	Stable
Features	Supports streaming response, thinking content

3. Google Generative AI Interface Recommended

Item	Details
Base URL	https://api.agentsflare.com/google/v1/models/
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	Gemini (Request method), Nano Banana, Veo
Service Status	Stable
Features	Supports automatic channel switching on 429, SSE streaming

4. BytePlus Interface (Seedance) Recommended

Item	Details
Base URL	https://api.agentsflare.com/byteplus/v3/contents/generations/tasks
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	Seedance
Service Status	Stable

5. Kling AI Interface Recommended

Item	Details
Base URL	https://api.agentsflare.com/klingai/v1/videos
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	Kling v2.5 Turbo, Kling v1.6
Service Status	Stable
Features	Supports text2video, image2video, multi-image2video

6. Zhipu OCR Interface Recommended

Item	Details
Base URL	https://api.agentsflare.com/zhipu/v4/layout_parsing
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	glm-ocr
Service Status	Stable
Features	Supports both image URL and base64 transmission; focused on OCR and layout parsing

7. MiniMax Video Generation Interface (MiniMax-H3) Recommended

Item	Details
Base URL	https://api.agentsflare.com/minimax/v2
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	MiniMax-H3
Service Status	Stable
Features	Text-to-video, image-to-video (first/last frame), reference-to-video, 768P/2K output, H3-Context-IR prompt enhancement

Endpoints

Endpoint	Method	Description
https://api.agentsflare.com/minimax/v2/video_generation	POST	Create a video generation task
https://api.agentsflare.com/minimax/v2/h3_context_ir	POST	Create an H3-Context-IR prompt enhancement task
https://api.agentsflare.com/minimax/v2/video_regeneration	POST	Create a video regeneration task (768P → 2K)
https://api.agentsflare.com/minimax/v2/query/video_generation/{task_id}	GET	Query task status/result (shared by all three task types)

See MiniMax-H3 API Examples for detailed request/response samples.

Legacy Interfaces (Maintenance Mode)

8. Legacy Google API Interface No Longer Updated

Item	Details
Base URL	https://api.agentsflare.com/googleapi/google/v1/models/
Authentication Method	Bearer Token
Content Type	application/json
Applicable Models	Gemini, Nano Banana, Veo
Service Status	Maintenance Mode
Note	Still available but no longer updated. Migration to new address recommended

Deployment Information

Item	Details
Deployment Location	Singapore
Network Latency	Approximately 100ms (low latency level)
Domestic Access Status	Does not affect normal access for domestic users
Availability	99.9% SLA

Address Selection Guide

💡 Selection Recommendations

1. Using OpenAI SDK? Use https://api.agentsflare.com/v1
2. Using Claude or MiniMax? Use https://api.agentsflare.com/anthropic/v1/messages
3. Using Google models (Request method)? Use https://api.agentsflare.com/google/v1/models/
4. Using Seedance? Use https://api.agentsflare.com/byteplus/v3/contents/generations/tasks
5. Using Kling AI? Use https://api.agentsflare.com/klingai/v1/videos
6. Using GLM-OCR? Use https://api.agentsflare.com/zhipu/v4/layout_parsing
7. Using MiniMax-H3 (video generation)? Use https://api.agentsflare.com/minimax/v2/video_generation (see Endpoints above for the full list)
8. Using text-embedding-v4 (embeddings)? Use https://api.agentsflare.com/v1/embeddings (same OpenAI-compatible base URL as chat completions)
9. If you're still using the old Google API address, migrate to the new address for better service

Feature Comparison

Feature	OpenAI SDK Interface	Anthropic Interface	Google New Interface	BytePlus Interface	Kling AI Interface	Zhipu OCR Interface	MiniMax-H3 Interface	Google Old Interface
OpenAI SDK Compatible	✅	❌	❌	❌	❌	❌	❌	❌
Anthropic SDK Compatible	❌	✅	❌	❌	❌	❌	❌	❌
429 Auto Channel Switch	✅	✅	✅	✅	✅	✅	✅	❌
Streaming Response	✅	✅	✅	❌	❌	❌	❌	✅
Continuous Updates	✅	✅	✅	✅	✅	✅	✅	❌
Recommended	✅	✅	✅	✅	✅	✅	✅	⚠️

FAQ

How to choose the right API address?

Choose based on the model and SDK you're using:

• Using OpenAI SDK? Choose https://api.agentsflare.com/v1
• Using Anthropic SDK? Choose https://api.agentsflare.com/anthropic/v1/messages
• Using Native Google API? Choose https://api.agentsflare.com/google/v1/models/
• Using Seedance? Choose https://api.agentsflare.com/byteplus/v3/contents/generations/tasks
• Using Kling AI? Choose https://api.agentsflare.com/klingai/v1/videos
• Using GLM-OCR? Choose https://api.agentsflare.com/zhipu/v4/layout_parsing
• Using MiniMax-H3 (video generation)? Choose https://api.agentsflare.com/minimax/v2/video_generation (create), /h3_context_ir (prompt enhancement), /video_regeneration (768P→2K regeneration), /query/video_generation/{id} (query)
• Using text-embedding-v4 (embeddings)? Choose https://api.agentsflare.com/v1/embeddings (OpenAI SDK compatible)

Can I still use the old address?

The old Google API address (https://api.agentsflare.com/googleapi/google/v1/models/) is still available but no longer receives feature updates and optimizations. We recommend migrating to the new address for better service quality.

Will migrating to the new address affect existing services?

No. Old and new addresses can coexist. You can migrate gradually without affecting existing services.