Choosing the right AI model
Anuma gives you access to 40+ AI models across leading providers — all through a single interface with one unified memory. Different models excel at different tasks, so choosing the right one can make a big difference.
Available models
Closed source models
| Provider | Models |
|---|---|
| OpenAI | GPT 5.5, ChatGPT 5.4, GPT 5.2 (standard-mode default), GPT-4o, o3, o4-mini |
| Anthropic | Claude Opus 4.7, Claude Sonnet 4.6 |
| Gemini 3.1 Pro, Gemini 3.1 Fast, Gemma 4 | |
| xAI | Grok 4.3, Grok 4.1 (Reasoning), Grok 4.1 (Non-Reasoning) |
| MiniMax | MiniMax 2.7, M2.5, M2.1 |
Open source models
| Provider | Models |
|---|---|
| DeepSeek | DeepSeek V4 Flash, DeepSeek V3.2 |
| Kimi | Kimi 2.6, Kimi 2.5 |
| Qwen | Qwen 3.6 Max Preview, Qwen 3.6 Plus, Qwen 3 235B |
| GLM/Zhipu | GLM-5, GLM 4.7 |
| InclusionAI | Ling 2.6 Flash |
Image models
| Provider | Models |
|---|---|
| Nano Banana | Nano Banana 2, Pro, Flash |
| Flux | Flux 1 Dev, Flux 1 Schnell |
How to select a model
Step 1: Open the model selector
Tap the model name at the top of any chat to open the model picker.
Step 2: Browse by provider
Models are grouped by provider. Scroll through or use search to find a specific model.
Step 3: Select your model
Every model is available on every plan — Free, Starter, and Pro. Tap a model to select it for your conversation.
Which model should I use?
- General chat and Q&A — Auto Model or GPT 5.2 (standard-mode default)
- Long documents and analysis — Claude Opus 4.7
- Coding and debugging — GPT 5.5 or Claude Sonnet 4.6
- Creative writing — Claude or Gemini
- Maximum privacy — DeepSeek, Kimi, Qwen, Ling 2.6 Flash, or any open source model
- Fast responses — Gemini Flash, Grok Non-Reasoning, or DeepSeek V4 Flash
Good to know
- All models share your unified Anuma memory. Switch between models without losing context.
- Different models cost different amounts of credits per message. Premium models cost more.
- Use Auto Model to let Anuma choose the best model automatically.
- You can mix models in Council Mode to compare responses side by side.
- The model list is continuously updated as new models become available — see What’s new for recent additions.
Related articles
Last updated on