Models
Simo models: Simo-1 and Simo-1 Pro
Two models, one API. Same question types, same calibrated probabilities. Choose by how much the decision matters.
At a glance
| Model | API id | For |
|---|---|---|
| Simo-1 | simo-1 | The default. Real-time judgments for agents, pipelines and per-event processing. |
| Simo-1 Pro | simo-1-pro | Maximum judgment quality for high-stakes gating and hard calls. |
Latency
Measured on one request over one 1280×720 screenshot, with the questions answered in the same pass.
| Model | 1 question | 10 questions | Per added question |
|---|---|---|---|
| Simo-1 | 64 ms | 104 ms | ~10 ms |
| Simo-1 Pro | 189 ms | 324 ms | ~32 ms |
Reading the situation is the cost; questions are nearly free. Ask everything you want to know in one request.
Which model should I use?
- Start with Simo-1 for the agent loop, the ticket stream, the per-event pipeline and anything where latency is felt.
- Use Simo-1 Pro when a wrong call is costly: gating an irreversible action, deciding a hard case, or pointing at the exact spot on screen.
- You can mix them: Simo-1 for every step, Simo-1 Pro for the few that matter.
Coming next
- The full agent step on Simo-1 Pro: decision, argument values, extraction and self-checks together, in under half a second.
- Fill and Extract graduating from preview, with a published evaluation.
- Arguments and extraction directly from screenshots and video.
Frequently asked questions
What is the difference between Simo-1 and Simo-1 Pro?
Both answer the same question types with calibrated probabilities. Simo-1 is the real-time default. Simo-1 Pro gives maximum judgment quality for high-stakes gating and hard calls, at a higher latency.
What are the API ids?
simo-1 and simo-1-pro.