StepFun

Step 3

Type
language model
Context
66K tokens
Max output
8K
Released
31 July 2025
Knowledge cutoff
January 2025
API string
step-3

Prices by provider

Provider Input, $ per 1M Output, $ per 1M In our data since
ZenMux $0.21 $0.57 31 Jul 2026
NanoGPT $0.25 $0.649 31 Jul 2026

Only the base tier and only per-token prices are shown. Batch, discounted and cached rates, as well as prices per image or per second of video, do not go into this table: they cannot stand in the same column as a price per million tokens.

The date in the last column is the day this price first entered our collection. The price may well be older: before that day we simply were not recording it. It has not changed since — otherwise a new row with a new date would stand in its place.

Measurement results

Task set Result Run conditions Measured by
Arena Score in Spanish 1,380.53 1,338–1,424 — —
Arena Score in English 1,370.04 1,360–1,380 — —
Arena Score, mathematics 1,367.15 1,336–1,398 — —
Arena Score, programming 1,365.68 1,349–1,382 — —
Arena Score, hard prompts 1,354.57 1,344–1,365 — —
Arena Score, overall 1,349.74 1,342–1,357 — —
Arena Score, multi-turn dialogue 1,338.08 1,320–1,356 — —
Arena Score, instruction following 1,332.15 1,318–1,347 — —
Arena Score, expert questions 1,330.51 1,296–1,365 — —
Arena Score, long queries 1,327.15 1,311–1,343 — —
Arena Score in Russian 1,324.81 1,297–1,353 — —
Arena Score, creative writing 1,318.79 1,298–1,340 — —
Arena Score, working with images 1,176.31 1,165–1,188 — —
Arena Score, text recognition in images 1,168.14 1,151–1,185 — —
Arena Score, understanding diagrams 1,153.08 1,123–1,183 — —