# Step 3.5 Flash 2603 API prices

Step 3.5 Flash 2603 — 262k context, 256k max output.
4 metered per-token offers from 6 provider(s), in USD per 1M tokens.

| Provider | Input $/1M | Output $/1M | Cache read $/1M | Context |
| --- | --- | --- | --- | --- |
| EmpirioLabs AI | $0.1 | $0.3 | $0.02 | 256k |
| NanoGPT | $0.1 | $0.3 | $0.05 | 262k |
| StepFun (China) | $0.1 | $0.3 | $0.02 | 256k |
| StepFun (Global) | $0.1 | $0.3 | $0.02 | 256k |

## Summary

- Cheapest input: $0.1 per 1M tokens (EmpirioLabs AI)
- Cheapest output: $0.3 per 1M tokens (EmpirioLabs AI)
- First-party: not listed separately
- Free offers: none
- Subscription plans (not per-token): StepFun Step Plan (Global), StepFun Step Plan (China)
- Released: 2026-04-02
- Inputs: text

HTML page: https://llmprice.gitlab.io/models/step-3-5-flash-2603/
Full dataset: https://llmprice.gitlab.io/data/catalog.json