VexicAI models in Vexic
Vexic has 18 models from different developers. You pick the model in the chat and can change it at any time: a light one for a simple task, a strong one for a hard task.
Works in the browser on your computer and phone.
All models
| Model | Understands images | Limit usage |
|---|---|---|
| GPT-6 Astra | yes | x5 |
| Claude Opus 5.5 | yes | x10 |
| Claude Fable 5.1 | yes | x24 |
| Claude Sonnet 5.5 | yes | x5 |
| Claude Haiku 5.5 | yes | x0.2 |
| Grok 4.7 | yes | x1 |
| Gemini 4 Argon | yes | x2.7 |
| Gemini 3.1 Pro | yes | x3 |
| GPT-6.1 Sol | yes | x1 |
| GPT-5.6 Sol | yes | x2.2 |
| GPT-5.6 Terra | yes | x1.1 |
| GPT-5.6 Luna | yes | x0.1 |
| Gemini 3.8 Flash | yes | x1 |
| DeepSeek V4 Pro | no, text only | x2 |
| DeepSeek V4 Flash | no, text only | x0.8 |
| GLM 5.3 | no, text only | x2 |
| Gemini 3.5 Flash Lite | yes | x0.5 |
| GPT-6 Luna | yes | x0.05 |
Limit usage shows how fast a model spends your limit compared with an x1 model. Stronger models use it faster. More on the pricing page.
By developer
GPT (OpenAI)
GPT-6 Astra, GPT-6.1 Sol, GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna, GPT-6 Luna
Claude (Anthropic)
Claude Opus 5.5, Claude Fable 5.1, Claude Sonnet 5.5, Claude Haiku 5.5
Gemini (Google)
Gemini 4 Argon, Gemini 3.1 Pro, Gemini 3.8 Flash, Gemini 3.5 Flash Lite
Grok (xAI)
Grok 4.7
DeepSeek (DeepSeek)
DeepSeek V4 Pro, DeepSeek V4 Flash
GLM (Zhipu AI)
GLM 5.3
How to choose a model
- For simple questions and short texts a light model with usage up to x1 is enough: it uses little of your limit and answers fast.
- For documents, spreadsheets and code mid-level models work well.
- For big and complex tasks, where care over a long run matters, choose strong models. They use the limit faster.
- If your task has images, pick a model that understands them (see the table).
The free plan includes GPT-6 Luna and Claude Haiku 5.5. All other models are available on any paid plan.
Frequently asked questions
Can I switch models in the same chat?
Yes. You choose the model before sending a message, and the next turn runs on the new model with the same conversation and files.
What happens if a model does not respond?
Vexic retries the request several times, and if the model still does not answer, it continues on another model with a similar price. You see a note about it in the chat.
What does usage x8 mean?
A model with usage x8 spends your limit about eight times faster than an x1 model on the same work.
How do I pay?
Paid plans are paid in rubles with a card issued by a Russian bank, through the YooKassa payment service. Cards issued in other countries are not accepted. A plan lasts 30 days and does not renew automatically: to extend it, you pay again. The Basic plan costs 2,000 ₽.
Try it on your own task
Sign in with your email, Yandex ID or VK ID. On the free plan you can give your first task right away.
Try for free