[Model] Add new model provider: AI/ML API

[D1m7asis]

From https://github.com/open-compass/VLMEvalKit/issues/1073

This PR adds first-class support for AI/ML API as a new unified model provider to VLMEvalKit. AI/ML API provides access to 300+ models, including GPT-4o, Claude 3.5, Gemini 1.5, DeepSeek, LLaMA3, Mistral, and many others — via a single OpenAI-compatible interface.

---

🔧 What was added

• ✅ New model wrapper: vlmeval/api/aimlapi.py
• ✅ Registration in vlmeval/api/__init__.py
• ✅ Multiple preconfigured model entries in vlmeval/config.py via ai_ml_api_models
• ✅ Extended api_models with support for AIMLAPI
• ✅ English and Chinese documentation in docs/en/Quickstart.md and docs/zh-CN/Quickstart.md
• ✅ Ready-to-use AIML_API_KEY environment variable support

---

🚀 Usage Example

AIML_API_KEY=your_key_here python run.py --data CCBench --model aimlapi_gpt_4o

---

✅ Checklist

• [x] Code follows repo structure
• [x] Docs (EN + ZH) updated
• [x] Tested on: MMBench, CCBench, MME
• [x] Uses OpenAI-compatible API structure

---

Let me know if you'd like to adjust or split this PR into parts.

[kennymckormick]

Hi, @D1m7asis

Would you please share me a temporary account on the AI / ML API platform for testing (w. $10 credit, e.g.)? My email is [email protected].

[D1m7asis]

@kennymckormick hi! Can i get a review here pls? Thanks!

[OctavianTheI]

Hi @kennymckormick - any chance we could get a review here? We'll implement changes within a day if anything needs to be adjusted