Skip to content

[docs] MaaS - add AI model concept overview and per-model modalities … #27

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
greennode-vdb merged 1 commit into from
Jul 31, 2026
Merged

[docs] MaaS - add AI model concept overview and per-model modalities … #27

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions English/SUMMARY.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -736,6 +736,7 @@
* [Import Model Registry Using vLLM](ai-stack/ai-platform/model-registry/import-model-registry-using-vllm.md)
* [Import Model Registry Using a Custom Container](ai-stack/ai-platform/model-registry/import-model-registry-using-a-custom-container.md)
* [Model as a Service (MaaS)](ai-stack/model-as-a-service/README.md)
* [Understanding AI Models](ai-stack/model-as-a-service/understanding-ai-models.md)
* [OCR API Usage Guide](ai-stack/model-as-a-service/ocr-api-usage-guide.md)
* [Playground](ai-stack/model-as-a-service/playground.md)
* [Available Models](ai-stack/model-as-a-service/available-models.md)
Expand Down
70 changes: 35 additions & 35 deletions English/ai-stack/model-as-a-service/available-models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,53 +12,53 @@ MaaS hỗ trợ nhiều loại mô hình khác nhau:

#### _**Chat**_

| Model Name | Provider | Rate Limit | Available From |
| --------------- | ----------- | ---------- | ------------------ |
| Kimi K2.7 Code | Moonshot AI | – | August 3, 2026 |
| Kimi 2.6 | Moonshot AI | – | August 3, 2026 |
| GLM-5.2 | Zhipu AI | – | August 3, 2026 |
| Gemma 4 31B-IT | Google | – | August 3, 2026 |
| Qwen 3.7 Plus | Qwen | – | August 3, 2026 |
| Qwen 3.6 Plus | Qwen | – | August 3, 2026 |
| Qwen 3.6 Flash | Qwen | – | August 3, 2026 |
| MiniMax M3 | MiniMax | – | August 3, 2026 |
| MiniMax M2.5 | MiniMax | – | August 3, 2026 |
| Opus 4.8 | Anthropic | – | August 20, 2026 |
| Sonnet 4.6 | Anthropic | – | August 20, 2026 |
| Haiku 4.5 | Anthropic | – | August 20, 2026 |
| GPT-5.4 | OpenAI | – | August 20, 2026 |
| GPT-5.4 Mini | OpenAI | – | August 20, 2026 |
| GPT-5.4 Nano | OpenAI | – | August 20, 2026 |
| GPT-5 | OpenAI | – | August 20, 2026 |
| GPT-5 Mini | OpenAI | – | August 20, 2026 |
| GPT-5 Nano | OpenAI | – | August 20, 2026 |
| GPT-4o | OpenAI | – | August 20, 2026 |
| GPT-4o Mini | OpenAI | – | August 20, 2026 |
| GPT-OSS 20B | OpenAI | – | August 20, 2026 |
| GPT-OSS 120B | OpenAI | – | August 20, 2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Available From |
| --------------- | ----------- | ----------------------------------- | ---------- | ------------------ |
| Kimi K2.7 Code | Moonshot AI | Text + Image → Text | – | August 3, 2026 |
| Kimi 2.6 | Moonshot AI | Text + Image + Video → Text | – | August 3, 2026 |
| GLM-5.2 | Zhipu AI | Text → Text | – | August 3, 2026 |
| Gemma 4 31B-IT | Google | Text + Image → Text | – | August 3, 2026 |
| Qwen 3.7 Plus | Qwen | Text + Image + Video → Text | – | August 3, 2026 |
| Qwen 3.6 Plus | Qwen | Text + Image + Video → Text | – | August 3, 2026 |
| Qwen 3.6 Flash | Qwen | Text + Image → Text | – | August 3, 2026 |
| MiniMax M3 | MiniMax | Text + Image + Video → Text | – | August 3, 2026 |
| MiniMax M2.5 | MiniMax | Text → Text | – | August 3, 2026 |
| Opus 4.8 | Anthropic | Text + Image → Text | – | August 20, 2026 |
| Sonnet 4.6 | Anthropic | Text + Image → Text | – | August 20, 2026 |
| Haiku 4.5 | Anthropic | Text + Image → Text | – | August 20, 2026 |
| GPT-5.4 | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-5.4 Mini | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-5.4 Nano | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-5 | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-5 Mini | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-5 Nano | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-4o | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-4o Mini | OpenAI | Text + Image → Text | – | August 20, 2026 |
| GPT-OSS 20B | OpenAI | Text → Text | – | August 20, 2026 |
| GPT-OSS 120B | OpenAI | Text → Text | – | August 20, 2026 |

{% hint style="info" %}
Models listed as **August 3, 2026** are live now. Models listed as **August 20, 2026** are rolling out progressively and are not yet available.
Modalities describe the model's general capability per its provider's documentation — actual support may depend on the specific API endpoint used via MaaS. Models listed as **August 3, 2026** are live now; models listed as **August 20, 2026** are rolling out progressively and are not yet available.
{% endhint %}

#### _Image Generation_

| Model Name | Provider | Rate Limit | Available From |
| -------------- | -------- | ---------- | --------------- |
| gpt-image-2 | OpenAI | – | August 20, 2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Available From |
| -------------- | -------- | ------------------------------- | ---------- | --------------- |
| gpt-image-2 | OpenAI | Text + Image → Image | – | August 20, 2026 |

#### _Embedding_

| Model Name | Provider | Rate Limit | Available From |
| -------------------- | -------- | ---------- | --------------- |
| Cohere Embed v4 | Cohere | – | August 20, 2026 |
| Qwen3 Embedding 8B | Qwen | – | August 20, 2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Available From |
| -------------------- | -------- | ------------------------------- | ---------- | --------------- |
| Cohere Embed v4 | Cohere | Text + Image → Vector | – | August 20, 2026 |
| Qwen3 Embedding 8B | Qwen | Text → Vector | – | August 20, 2026 |

#### _Rerank_

| Model Name | Provider | Rate Limit | Available From |
| ------------------------ | -------- | ---------- | --------------- |
| Cohere Rerank v4.0 pro | Cohere | – | August 20, 2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Available From |
| ------------------------ | -------- | ------------------------------- | ---------- | --------------- |
| Cohere Rerank v4.0 pro | Cohere | Text → Ranked List | – | August 20, 2026 |

Note:

Expand Down
63 changes: 63 additions & 0 deletions English/ai-stack/model-as-a-service/understanding-ai-models.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,63 @@
---
description: >-
A functional overview of AI models on MaaS — purpose of use, basic principle
of operation, and the main input data type for each model category.
---

# Understanding AI Models

GreenNode MaaS lets you use leading AI models right away — for chat, coding, image understanding, and document processing — without training or operating your own GPU infrastructure. This page explains what AI models are for, how they work at a basic level, and what kind of input data each model category expects, so you can pick the right model before browsing [Available Models](available-models.md).

***

## Purpose of Use

AI models on MaaS are pre-trained deep learning models that automate tasks previously requiring direct human effort: answering questions, summarizing or drafting content, generating and fixing code, reading images, generating images from a description, extracting data from documents, or searching by meaning. Instead of writing hard-coded rules for every case, you describe what you need in natural language (a prompt), and the model reasons out a suitable result.

## Principle of Operation

At a basic functional level, an AI model processes a request in three steps:

```
Input (text / image / audio / PDF)
│
▼
Tokenize the input data
│
▼
Model predicts/generates output based on
patterns learned from training data
│
▼
Output (text / image / vector / JSON...)
```

* **Tokenization:** Input data (words, image pixels...) is broken down into small units called **tokens**.
* **Context-based prediction:** A model doesn't "understand" the way a human does — it predicts the most likely next token based on the full context (the prompt, conversation history, system prompt) and patterns learned from a massive amount of training data.
* **Thinking vs. non-thinking modes:** Some models — especially "Reasoning" models such as Qwen, GLM, and DeepSeek — support a **thinking** mode (step-by-step, chain-of-thought reasoning) that improves accuracy on multi-step problems at the cost of a slower response; **non-thinking** mode answers faster for simpler tasks.

{% hint style="info" %}
The more tokens in context (prompt + conversation history), the longer a model takes to process and the higher the cost — see how token-based billing works in [Pricing](pricing.md).
{% endhint %}

## Model Categories by Function

| Category | Purpose of Use | Main Input Data | Typical Use-cases |
| ----------------------------- | ------------------------------------------------------------------------------ | ---------------------------------------- | -------------------------------------------------------------------------- |
| Chat / Drafting | Q\&A, summarization, writing & editing content | Text, system prompt | Virtual assistants, report drafting, internal Q\&A |
| Code | Generate/fix code, code review, write unit tests | Text/code, multi-file context | Script automation, multi-file edits, code-run-fix loops |
| Reasoning | Solve multi-step logic/business problems, plan, check constraints | Text describing the problem/constraints | Multi-step calculations, constraint checking, planning |
| Vision (multimodal) | Read, describe, and classify image content | Image + text | Reading charts/screenshots, UI checks, extracting content from images |
| Image Generation | Generate new images from a text description | Text (prompt) | Banners, illustrations, mockups for social/marketing |
| OCR / Document AI (IDP) | Extract text and structured fields from documents | Scanned images or PDF files | Invoice extraction, ID/passport/driver's license reading, image-to-text |
| Embedding (RAG · Step 1) | Generate a semantic vector representation of text | Text | Semantic search, building RAG pipelines, internal document search |
| Rerank (RAG · Step 2) | Re-rank retrieved results by relevance | (Query, retrieved passage) pairs | Improving accuracy/relevance for RAG and search systems |

## Getting Started

| I want to... | Go to |
| ---------------------------------------------- | ------------------------------------------ |
| See models grouped by function | [Available Models](available-models.md) |
| See unit prices by model | [Model Pricing List](model-pricing-list.md) |
| Quickly try a model before integrating | [Playground](playground.md) |
| Call a model via API | [MaaS API](maas-api.md) |
1 change: 1 addition & 0 deletions Vietnamese/SUMMARY.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -825,6 +825,7 @@
* [Import Model Registry sử dụng vLLM](ai-stack/ai-platform/model-registry/import-vllm-model.md)
* [Import Model Registry với custom container](ai-stack/ai-platform/model-registry/import-model-registry-voi-custom-container.md)
* [Model as a Service (MaaS)](ai-stack/model-as-a-service/README.md)
* [Giới thiệu về AI Model](ai-stack/model-as-a-service/gioi-thieu-ve-ai-model.md)
* [Hướng dẫn sử dụng API OCR](ai-stack/model-as-a-service/huong-dan-su-dung-api-ocr.md)
* [Playground](ai-stack/model-as-a-service/playground.md)
* [Các Model được cung cấp](ai-stack/model-as-a-service/cac-model-duoc-cung-cap.md)
Expand Down
70 changes: 35 additions & 35 deletions Vietnamese/ai-stack/model-as-a-service/cac-model-duoc-cung-cap.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -12,53 +12,53 @@ MaaS hỗ trợ nhiều loại mô hình khác nhau:

#### _Chat_

| Model Name | Provider | Rate Limit | Sẵn có từ |
| --------------- | ----------- | ---------- | ------------ |
| Kimi K2.7 Code | Moonshot AI | – | 03/08/2026 |
| Kimi 2.6 | Moonshot AI | – | 03/08/2026 |
| GLM-5.2 | Zhipu AI | – | 03/08/2026 |
| Gemma 4 31B-IT | Google | – | 03/08/2026 |
| Qwen 3.7 Plus | Qwen | – | 03/08/2026 |
| Qwen 3.6 Plus | Qwen | – | 03/08/2026 |
| Qwen 3.6 Flash | Qwen | – | 03/08/2026 |
| MiniMax M3 | MiniMax | – | 03/08/2026 |
| MiniMax M2.5 | MiniMax | – | 03/08/2026 |
| Opus 4.8 | Anthropic | – | 20/08/2026 |
| Sonnet 4.6 | Anthropic | – | 20/08/2026 |
| Haiku 4.5 | Anthropic | – | 20/08/2026 |
| GPT-5.4 | OpenAI | – | 20/08/2026 |
| GPT-5.4 Mini | OpenAI | – | 20/08/2026 |
| GPT-5.4 Nano | OpenAI | – | 20/08/2026 |
| GPT-5 | OpenAI | – | 20/08/2026 |
| GPT-5 Mini | OpenAI | – | 20/08/2026 |
| GPT-5 Nano | OpenAI | – | 20/08/2026 |
| GPT-4o | OpenAI | – | 20/08/2026 |
| GPT-4o Mini | OpenAI | – | 20/08/2026 |
| GPT-OSS 20B | OpenAI | – | 20/08/2026 |
| GPT-OSS 120B | OpenAI | – | 20/08/2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Sẵn có từ |
| --------------- | ----------- | ---------------------------------- | ---------- | ------------ |
| Kimi K2.7 Code | Moonshot AI | Text + Image → Text | – | 03/08/2026 |
| Kimi 2.6 | Moonshot AI | Text + Image + Video → Text | – | 03/08/2026 |
| GLM-5.2 | Zhipu AI | Text → Text | – | 03/08/2026 |
| Gemma 4 31B-IT | Google | Text + Image → Text | – | 03/08/2026 |
| Qwen 3.7 Plus | Qwen | Text + Image + Video → Text | – | 03/08/2026 |
| Qwen 3.6 Plus | Qwen | Text + Image + Video → Text | – | 03/08/2026 |
| Qwen 3.6 Flash | Qwen | Text + Image → Text | – | 03/08/2026 |
| MiniMax M3 | MiniMax | Text + Image + Video → Text | – | 03/08/2026 |
| MiniMax M2.5 | MiniMax | Text → Text | – | 03/08/2026 |
| Opus 4.8 | Anthropic | Text + Image → Text | – | 20/08/2026 |
| Sonnet 4.6 | Anthropic | Text + Image → Text | – | 20/08/2026 |
| Haiku 4.5 | Anthropic | Text + Image → Text | – | 20/08/2026 |
| GPT-5.4 | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-5.4 Mini | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-5.4 Nano | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-5 | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-5 Mini | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-5 Nano | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-4o | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-4o Mini | OpenAI | Text + Image → Text | – | 20/08/2026 |
| GPT-OSS 20B | OpenAI | Text → Text | – | 20/08/2026 |
| GPT-OSS 120B | OpenAI | Text → Text | – | 20/08/2026 |

{% hint style="info" %}
Model có mốc **03/08/2026** đã sẵn sàng sử dụng. Model có mốc **20/08/2026** đang được triển khai dần và chưa khả dụng.
Modalities mô tả khả năng chung của model theo tài liệu của nhà cung cấp — khả năng thực tế có thể phụ thuộc vào endpoint API cụ thể được gọi qua MaaS. Model có mốc **03/08/2026** đã sẵn sàng sử dụng; model có mốc **20/08/2026** đang được triển khai dần và chưa khả dụng.
{% endhint %}

#### _Image Generation_

| Model Name | Provider | Rate Limit | Sẵn có từ |
| -------------- | -------- | ---------- | ---------- |
| gpt-image-2 | OpenAI | – | 20/08/2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Sẵn có từ |
| -------------- | -------- | ------------------------------- | ---------- | ---------- |
| gpt-image-2 | OpenAI | Text + Image → Image | – | 20/08/2026 |

#### _Embedding_

| Model Name | Provider | Rate Limit | Sẵn có từ |
| -------------------- | -------- | ---------- | ---------- |
| Cohere Embed v4 | Cohere | – | 20/08/2026 |
| Qwen3 Embedding 8B | Qwen | – | 20/08/2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Sẵn có từ |
| -------------------- | -------- | ------------------------------- | ---------- | ---------- |
| Cohere Embed v4 | Cohere | Text + Image → Vector | – | 20/08/2026 |
| Qwen3 Embedding 8B | Qwen | Text → Vector | – | 20/08/2026 |

#### _Rerank_

| Model Name | Provider | Rate Limit | Sẵn có từ |
| ------------------------ | -------- | ---------- | ---------- |
| Cohere Rerank v4.0 pro | Cohere | – | 20/08/2026 |
| Model Name | Provider | Modalities (Input → Output) | Rate Limit | Sẵn có từ |
| ------------------------ | -------- | ------------------------------- | ---------- | ---------- |
| Cohere Rerank v4.0 pro | Cohere | Text → Ranked List | – | 20/08/2026 |

Lưu ý:

Expand Down
Loading
Loading