---
title: "Model Catalog"
description: "Catalog of AI models available on AINative Studio across text, code, reasoning, and media"
canonical: "https://docs.ainative.studio/docs/api/models"
last-updated: "2026-10-03T21:47:10.764Z"
---

# Model Catalog

Source: https://docs.ainative.studio/docs/api/models

> Catalog of AI models available on AINative Studio across text, code, reasoning, and media

# Model Catalog

AINative Studio provides access to models across text, code, reasoning, image, video, audio, and embedding categories. Models are organized into four access tiers:

| Tier | Access |
|------|--------|
| **Free** | Available to all users, no subscription required |
| **Basic** | Requires Basic plan ($9/mo) |
| **Professional** | Requires Professional plan ($29/mo) |
| **Enterprise** | Requires Enterprise plan ($99/mo) |

All models are accessible through the unified `/v1/chat/completions` endpoint (text/code/reasoning) or their respective modality endpoints (image, video, audio, embeddings).

:::info Snapshot — the live list is authoritative
The tables below are a **snapshot** (last synced 2026-09-02). The authoritative, always-current list is the API itself: call **`GET /v1/public/models`** (see [Programmatic Access](#programmatic-access) below). If a model ID here disagrees with that endpoint, the endpoint wins. Deprecated aliases return an honest `model_deprecated` 400 that names a live alternative — they are not listed here as usable.
:::

---

## Text / Chat Models

General-purpose language models for conversation, content generation, and tool calling.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `llama-3.1-8b` | Llama 3.1 8B | NVIDIA NIM | Free | text, fast |
| `llama-4-maverick-17b-128e` | Llama 4 Maverick 17B | DigitalOcean | Free | text, multimodal |
| `llama-4-scout` | Llama 4 Scout | HuggingFace | Free | text, fast |
| `gpt-oss-120b` | GPT OSS 120B | Cerebras | Free | coding, tools, text, fast |
| `glm-5` | GLM-5 | DigitalOcean | Free | text, coding, reasoning |
| `glm-5.1` | GLM-5.1 | DigitalOcean | Professional | text, coding, reasoning |
| `glm-5.2` | GLM-5.2 | DigitalOcean | Professional | text, coding, reasoning |
| `glm-5.3` | GLM-5.3 (recommended) | DigitalOcean | Professional | text, coding, reasoning |
| `kimi-k3` | Kimi K3 | DigitalOcean | Professional | text, coding, reasoning |
| `deepseek-v4-pro` | DeepSeek V4 Pro | DigitalOcean | Professional | text, coding, reasoning |
| `claude-sonnet-4.5` | Claude Sonnet 4.5 | Anthropic | Professional | text, tools |
| `claude-sonnet-4.6` | Claude Sonnet 4.6 | Anthropic | Professional | text, tools |
| `claude-opus-4.5` | Claude Opus 4.5 | Anthropic | Enterprise | text, tools |
| `claude-opus-4.6` | Claude Opus 4.6 | Anthropic | Enterprise | text, tools |

---

## Code Models

Specialized for code generation, completion, and software engineering tasks.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `qwen-coder-7b` | Qwen Coder 7B | HuggingFace | Basic | coding |
| `qwen-coder-32b` | Qwen Coder 32B | HuggingFace | Basic | coding, tools |
| `qwen3-coder-30b-a3b` | Qwen3 Coder 30B A3B | HuggingFace | Professional | coding |
| `qwen3-coder-next` | Qwen3 Coder Next | HuggingFace | Professional | coding |
| `gpt-5.3-codex` | GPT-5.3 Codex | DigitalOcean | Professional | coding, fast |
| `minimax-m2.7` | MiniMax M2.7 | NVIDIA NIM | Professional | coding, reasoning |
| `qwen3.5-397b` | Qwen3.5 397B MoE | DigitalOcean | Professional | coding, tools, reasoning, fast |
| `deepseek-4-flash` | DeepSeek 4 Flash | DigitalOcean | Professional | text, fast |
| `kimi-k2` | Kimi K2 | DigitalOcean | Professional | text, coding, reasoning |

---

## Reasoning Models

Deep reasoning and chain-of-thought models for complex problem solving.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `deepseek-v3` | DeepSeek V3 | NVIDIA NIM | Professional | reasoning, text |
| `deepseek-3.2` | DeepSeek 3.2 | DigitalOcean | Professional | text, reasoning |
| `deepseek-r1-distill-qwen-7b` | DeepSeek R1 Distill Qwen 7B | NVIDIA NIM | Professional | reasoning |
| `deepseek-r1-distill-llama-8b` | DeepSeek R1 Distill Llama 8B | NVIDIA NIM | Professional | reasoning |
| `deepseek-r1-distill-llama-70b` | DeepSeek R1 Distill Llama 70B | DigitalOcean | Professional | reasoning |
| `deepseek-r1` | DeepSeek R1 | DigitalOcean | Enterprise | reasoning |
| `minimax-m2.5` | MiniMax M2.5 | DigitalOcean | Professional | text, coding, reasoning |
| `nemotron-3-super-120b` | Nemotron 3 Super 120B | NVIDIA | Professional | text, reasoning |
| `nemotron-3-ultra-550b` | Nemotron 3 Ultra 550B | NVIDIA | Enterprise | text, reasoning |

---

## Image Models

Text-to-image and image editing models.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `flux-schnell` | FLUX Schnell | HuggingFace | Professional | image |
| `flux-1-dev` | FLUX 1 Dev | HuggingFace | Professional | image |
| `flux-2` | Flux 2 | fal.ai | Professional | image |
| `seedream-5` | Seedream 5 | fal.ai | Professional | image |
| `nano-banana-2` | Nano Banana 2 | fal.ai | Professional | image |
| `grok-imagine` | Grok Imagine | fal.ai | Professional | image |
| `happy-horse` | Happy Horse | fal.ai | Professional | image |
| `sdxl` | Stable Diffusion XL | HuggingFace | Professional | image |
| `qwen-image-edit-2511-lora` | Qwen Image Edit | RunPod | Professional | image, edit |
| `sd-3.5-large` | Stable Diffusion 3.5 | HuggingFace | Enterprise | image |

---

## Video Models

Text-to-video and image-to-video generation models.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `kling-3-pro` | Kling 3 Pro | fal.ai | Enterprise | video |
| `veo-3` | Veo 3 | fal.ai | Enterprise | video |
| `runway-gen-4.5` | Runway Gen 4.5 | RunPod | Enterprise | video, premium |
| `hunyuanvideo-1.5` | HunyuanVideo 1.5 | HuggingFace | Enterprise | video |
| `cogvideox-2b` | CogVideoX 2B | HuggingFace | Enterprise | video |
| `seedance-v1-5-pro-i2v` | Seedance Pro i2v | RunPod | Enterprise | video |
| `sora-2-i2v` | Sora 2 i2v | RunPod | Enterprise | video |
| `wan-2-6-t2v` | Wan 2.6 T2V | RunPod | Enterprise | video |

---

## Audio / TTS Models

Speech-to-text (transcription) and text-to-speech models.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `whisper-small` | Whisper Small | HuggingFace | Basic | audio, transcription |
| `whisper-large-v3-turbo` | Whisper Large v3 Turbo | HuggingFace | Basic | audio, transcription |
| `minimax-speech-02-hd` | MiniMax Speech 02 HD | RunPod | Professional | audio, tts |

---

## Embedding Models

Vector embedding models for semantic search and RAG pipelines.

| Model ID | Label | Provider | Tier | Tags |
|----------|-------|----------|------|------|
| `all-minilm-l6-v2` | All MiniLM L6 v2 | HuggingFace | Free | embeddings |
| `bge-large-en` | BGE Large EN | HuggingFace | Basic | embeddings |

---

## Deprecated Model IDs

These aliases have been retired. A request for one returns a `model_deprecated` 400 that names a live alternative — they are **not** silently substituted. Update your integration to the suggested replacement.

| Deprecated ID | Use instead |
|---------------|-------------|
| `zai-glm-4.7`, `glm-4.7` | `glm-5.3` (or `glm-5` / `glm-5.1` / `glm-5.2`) |
| `qwen3-coder-flash` | `gpt-5.3-codex` or `qwen-coder-32b` |
| `claude-3-5-haiku` | `claude-sonnet-4.5` or `gpt-oss-120b` |
| `qwen-7b` | `qwen-coder-7b` or `glm-5` |
| `gpt-4-7550` | `gpt-4` (was a mislabeled catalog duplicate, not a distinct model) |
| `llama-3.3-8b-instruct` | `llama-3.1-8b` |
| `llama-3.3-70b-instruct`, `llama-3.3-70b` | `llama-4-maverick-17b-128e` (DigitalOcean) or `llama-4-scout` (HuggingFace) — removed from DigitalOcean's serverless catalog, and NVIDIA NIM reached end-of-life for this model 2026-08-26 |
| `qwq-32b`, `devstral`, `nouscoder-14b`, `nous-coder` | see `GET /v1/public/models` for current coding/reasoning models |

---

## Programmatic Access

Retrieve the full model catalog programmatically:

```bash
curl -X GET https://api.ainative.studio/v1/public/models \
  -H "X-API-Key: your-api-key"
```

To get only models available for your current subscription tier:

```bash
curl -X GET https://api.ainative.studio/v1/public/models/available \
  -H "X-API-Key: your-api-key"
```

Both endpoints return JSON with the following structure:

```json
{
  "ok": true,
  "models": [
    {
      "id": "llama-3.1-8b",
      "provider": "meta",
      "label": "Llama 3.1 8B",
      "capabilities": {
        "intelligence": 3,
        "speed": 5,
        "cost": 5
      },
      "context_window": 131072,
      "tags": ["text", "fast"],
      "tier": "free",
      "pricing": {
        "input": 0.0,
        "output": 0.0,
        "cache_read": 0.0
      }
    }
  ]
}
```

The response includes `ETag` and `Last-Modified` headers for cache validation.

---

## Pricing

Pricing is per 1M tokens (input/output) for text models, or per generation for image/video/audio models. Free-tier models have zero cost. See the [Billing documentation](/docs/billing/credits) for details on credit consumption.

Refs #3797