Mirai 1
Satuapps
Input
$0.10
per 1M tokens
Output
$0.80
per 1M tokens
- Context
- 200K
- Max output
- 33K
- Tool calling
- Image input
- Reasoning
Model id: mirai-1
the shelf, live
Every model you can call, and who builds it.
One key, one endpoint, one bill. Prices below are the ones the API charges from, read at render time — not a table someone remembered to update.
Featured
Mirai 2
Satuapps
Input
$0.50
per 1M tokens
Output
$1.50
per 1M tokens
- Context
- 500K
- Max output
- 131K
- Tool calling
- Image input
- Reasoning
Model id: mirai-2
What answers it
This brand routes each turn to the model that fits it. These are the ones it routes to today, and they change from the panel without a deploy.
DeepSeek V4 Pro
DeepSeek
Handles: chat
Kimi K3
Moonshot AI
Handles: code
GLM-5.3
Z.ai
Handles: analysis
Gemini 3.8 Flash
Google
Handles: images
Gemini 3.8 Flash
Input
$0.75
per 1M tokens
Output
$3.75
per 1M tokens
- Context
- 1.0M
- Max output
- 4K
- Tool calling
- Image input
- Reasoning
Model id: gemini-3.8-flash
Published scores
- LVBench87.8%vendor reported · 2026-09source
- Terminal-Bench 2.189.4%vendor reported · 2026-09source
- HLE-Verified54.9%vendor reported · 2026-09source
Each figure comes from an official publication, with its link and the month it appeared. Nothing here is averaged across models.