Step 3.5 Flash

stepfun· 2026-01-29· 262K ctx· apache-2.0

open weightsreasoning quant: unknown

spread (output)

1.0×

$0.300 – $0.300 · 16 providers

16 providers priced

Provider	Input	Output	Cache rd	Blended
Nvidia relay router	free/—	free/—	—	free/—
EmpirioLabs AI relay router	$0.100	$0.300	$0.020	$0.067
EmpirioLabs AI relay router	$0.100	$0.300	$0.020	$0.067
Hugging Face relay router	$0.100	$0.300	—	$0.118^*
Kilo Gateway relay router	$0.100	$0.300	—	$0.118^*
NanoGPT relay router	$0.100	$0.300	$0.050	$0.086
NanoGPT relay router	$0.100	$0.300	$0.050	$0.086
OpenRouter unknown	$0.100	$0.300	—	$0.118^*
SiliconFlow relay router	$0.100	$0.300	—	$0.118^*
SiliconFlow (China) relay router	$0.100	$0.300	—	$0.118^*
StepFun (China) first party lab	$0.100	$0.300	$0.020	$0.067
StepFun (China) first party lab	$0.100	$0.300	$0.020	$0.067
StepFun (Global) relay router	$0.100	$0.300	$0.020	$0.067
StepFun (Global) relay router	$0.100	$0.300	$0.020	$0.067
Vercel AI Gateway unknown	$0.090	$0.300	$0.020	$0.065
Eden AI relay router	$0.090	$0.300	$0.020	$0.065
ZenMux relay router	$0.100	$0.300	—	$0.118^*

Cheapest (output)

EmpirioLabs AI

$0.300 /M

opencode.json

{
  "provider": {
    "empiriolabs": {
      "npm": "@ai-sdk/openai-compatible",
      "options": {
        "baseURL": "https://api.empiriolabs.ai/v1"
      },
      "models": {
        "step-3-5-flash": {}
      }
    }
  }
}

Set EMPIRIOLABS_API_KEY in your environment. Generated — verify against the OpenCode docs.

Output $/M over time

Median line, min–max band. Pre-2026-09 points are catalog-recorded dates from models.dev git history, not provider effective dates.

Data: models.dev (MIT), snapshot 2026-09-20. Prices are vendor-published; verify before purchase. Methodology.