Supported Models
ModelMax routes requests to the best available provider. All models are accessed through a single API.
Chat models
Text-in, text-out conversation models. Used with POST /v1/chat/completions.
Google Gemini
| Model ID | Input | Output | Notes |
|---|
gemini-3.1-pro-preview | text, image, audio, video | text | Latest Gemini Pro |
gemini-3-pro-preview | text, image, audio, video | text | Gemini 3.0 Pro |
gemini-3-flash-preview | text, image, audio, video | text | Fast, cost-effective |
gemini-3.1-flash-image-preview | text, image | text, image | Image generation capable |
gemini-3.1-flash-lite-preview | text, image, audio, video | text | Lightweight |
xAI Grok on Google Cloud
| Model ID | Input | Output | Notes |
|---|
grok-4.3 | text, image | text | xAI flagship model hosted on Google Cloud |
grok-4.1-fast-non-reasoning | text, image | text | Fast, cost-effective non-reasoning model |
OpenAI
| Model ID | Input | Output | Notes |
|---|
gpt-5.5 | text, image | text | Flagship GPT model |
gpt-5.4 | text, image | text | High-capability GPT model |
gpt-5.4-mini | text, image | text | Balanced GPT model |
gpt-5.4-nano | text, image | text | Fast, low-cost GPT model |
Anthropic Claude
| Model ID | Input | Output | Notes |
|---|
claude-opus-4-8 | text, image | text | Highest-capability Claude model |
claude-sonnet-4-6 | text, image | text | Balanced Claude model |
claude-haiku-4-5 | text, image | text | Fast Claude model |
DeepSeek
| Model ID | Input | Output | Notes |
|---|
deepseek-r1 | text | text | Reasoning model with chain-of-thought |
deepseek-v3.1 | text | text | General purpose |
deepseek-v3.2 | text | text | Latest general purpose |
deepseek-v4-flash | text | text | Fast direct DeepSeek model |
deepseek-v4-pro | text | text | Higher-capability direct DeepSeek model |
Qwen
| Model ID | Input | Output | Notes |
|---|
qwen3-coder-30b-a3b | text | text | Code-focused, lightweight |
qwen3-32b | text | text | General purpose |
qwen3-235b-a22b-2507 | text | text | Large, capable |
qwen3-coder-480b-a35b | text | text | Code-focused, high capacity |
qwen3-next-80b-a3b | text | text | Efficient MoE architecture |
qwen3-vl-235b-a22b | text, image | text | Vision-language model |
qwen3-coder-next | text | text | Latest coder model |
qwen3-max | text | text | Flagship direct Qwen model |
qwen3.5-plus | text | text | Balanced direct Qwen model |
qwen3.5-flash | text | text | Fast direct Qwen model |
MiniMax
| Model ID | Input | Output | Notes |
|---|
minimax-m2 | text | text | MiniMax M2 |
minimax-m2.1 | text | text | MiniMax M2.1 |
minimax-m2.7 | text | text | Direct MiniMax model |
minimax-m2.5 | text | text | Direct MiniMax model |
Kimi (Moonshot)
| Model ID | Input | Output | Notes |
|---|
kimi-k2.6 | text | text | Latest direct Kimi model |
kimi-k2-thinking | text | text | With reasoning |
kimi-k2.5 | text | text | Latest Kimi |
Zhipu GLM
| Model ID | Input | Output | Notes |
|---|
glm-5.1 | text | text | Direct GLM model |
Volcengine Ark — Doubao
| Model ID | Input | Output | Notes |
|---|
doubao-seed-2-0-lite-260428 | text | text | Fast, efficient Doubao model |
doubao-seed-2-0-mini-260428 | text | text | Compact Doubao model |
doubao-seed-2-0-pro-260215 | text | text | Higher-capability Doubao model |
doubao-seed-2-0-code-preview-260215 | text | text | Code-focused Doubao model |
Baidu Qianfan — ERNIE
| Model ID | Input | Output | Notes |
|---|
ernie-5.1 | text | text | Direct ERNIE model |
Tencent TokenHub — Hunyuan
| Model ID | Input | Output | Notes |
|---|
hy3-preview | text | text | Hunyuan 3 preview model |
Pricing defaults are stored in the model seed data and can be adjusted in the admin model configuration when your provider contract differs from the public pricing.
Video models
Async video generation. Used with the Queue API (POST /v1/queue/{model}).
Google Gemini — Veo
| Model ID | Speed | Quality | Audio |
|---|
veo-3.1 | Standard | Highest | Yes |
veo-3.1-fast | Fast | High | Yes |
veo-3 | Standard | High | Yes |
veo-3-fast | Fast | Good | Yes |
veo-2 | Standard | Good | Yes |
All Veo models support:
- Text-to-video and image-to-video
- Up to 8 seconds duration
- 720p, 1080p, and 4K resolution
- Audio generation
Capability matrix
| Capability | Gemini Chat | Domestic Text | Qwen VL | Veo Video |
|---|
| Text input | Yes | Yes | Yes | Yes |
| Image input | Yes | — | Yes | Yes |
| Audio input | Yes | — | — | — |
| Video input | Yes | — | — | — |
| Text output | Yes | Yes | Yes | — |
| Image output | Partial | — | — | — |
| Video output | — | — | — | Yes |
| Streaming | Yes | Yes | Yes | — |
Choosing a model
For general chat: Start with gemini-3-flash-preview (fast, multimodal) or deepseek-v3.2 (strong text reasoning).
For code: Try qwen3-coder-480b-a35b or qwen3-coder-next.
For reasoning: Use deepseek-r1, glm-5.1, or kimi-k2.6.
For high-throughput domestic routing: Use doubao-seed-2-0-lite-260428, qwen3.5-flash, or hy3-preview.
For vision: Use gemini-3-flash-preview (images, audio, video input) or qwen3-vl-235b-a22b (images only).
For video generation: Use veo-3 for quality, veo-3-fast for speed.
