Models
Every model is available through the API with the id in the table. The chat offers the models that can read images and use tools.
Language models
| Model | Context | Tools | Vision | Reasoning | Input / 1M | Output / 1M |
|---|---|---|---|---|---|---|
Gemma 4 26Bgemma4-26b | 256.000 | ✓ | ✓ | ✓ | € 0,11 | € 0,55 |
Gemma4 Chat 26Bgemma4-chat-26b | 256.000 | ✓ | – | – | € 0,11 | € 0,55 |
GPT-OSS 120Bgpt-oss-120b | 128.000 | ✓ | – | ✓ | € 0,165 | € 0,605 |
Kolibri 1 TL5kolibri-1-tl5 | 262.144 | ✓ | – | ✓ | € 0,165 | € 0,605 |
Qwen 3.8 27Bqwen3.8-27b | 262.144 | ✓ | ✓ | ✓ | € 0,165 | € 0,715 |
Mistral 4 119Bmistral4-119b | 262.144 | ✓ | ✓ | ✓ | € 0,22 | € 0,66 |
Deepseek V4.1 Flash 1017 TL5deepseek-v4.1-flash-1017-tl5 | 1.048.576 | ✓ | ✓ | ✓ | € 0,275 | € 1,10 |
Deepseek V4.1 Flash 552Bdeepseek-v4.1-flash-552b | 1.048.576 | ✓ | ✓ | ✓ | € 0,275 | € 1,10 |
GLM 5.3 754Bglm5.3-754b | 524.288 | ✓ | – | ✓ | € 1,10 | € 3,85 |
GLM 5.3 TL5glm5.3-tl5 | 524.288 | ✓ | – | ✓ | € 1,10 | € 3,85 |
- Tools: the model can call functions you define (
tools). - Vision: the model reads images in a message.
- Reasoning: the model thinks before it answers; that thinking counts as output tokens.
Image models
| Model | Per megapixel |
|---|---|
Flux.2 [klein]flux2-klein | € 0,0055 |
Z-Image Turboz-image-turbo | € 0,0099 |
Qwen Imageqwen-image | € 0,0825 |
Live list
GET https://dev.typno.ai/v1/models returns the current list with context size and capabilities.
curl https://dev.typno.ai/v1/models -H "Authorization: Bearer $TYPNO_API_KEY"
The catalogue follows our provider. A model that is withdrawn there disappears from the list; requests for it then get model_not_found.