Skip to content
ModelCap

Launch report

Step 5 Preview launch report

ModelCap · generated from recorded data

StepFun's Step 5 Preview was first listed on 8 October 2026. It entered the ModelCap Index at #39 of 344 ranked language models on 9 October 2026, one day after its listed release date, backed by 3 admitted benchmarks: Arena Coding, Arena Overall and LMArena Agent. Its first measurement, on 8 October 2026 on one benchmark, Arena Overall, had placed it at #35 of 340; ModelCap counts a model as entered once two benchmarks back its position. As of 19 November 2026 it is #49 on the current board.

Position at entry · 9 October 2026
#39 of 344
Listing to first measurement
Same day
Position now · 19 November 2026
#49

Position at entry

Entry is the first board position backed by at least two admitted benchmarks. Read from the board's last published snapshot of each day named below; these values are fixed and do not change with later refreshes.

Listed release date
8 October 2026
First measurement · 8 October 2026
#35 of 340 ranked language modelsBacked by Arena Overall
Entry · 9 October 2026
#39 of 344 ranked language modelsBacked by Arena Coding, Arena Overall, LMArena Agent
Neighbours at entry
Directly above: GPT-5.4 MiniDirectly below: K2 Horizon 375B A23B

Benchmark results

Published results admitted to the Index, as of 19 November 2026. Each board links to ModelCap's page for it and each result to its source.

Current public benchmark results for Step 5 Preview
BenchmarkResultBoard positionPublished
Arena Overallgeneral14491442–1455#85 of 42218 November 2026
Arena Codingcoding15041493–1516#73 of 41718 November 2026
LMArena Agentagent-5.1%-6.6%–-3.7%#50 of 6818 November 2026

How this report is produced

ModelCap publishes a launch report when a newly released model first holds a measured position backed by at least two public benchmarks. The report is assembled automatically from ModelCap's own records: the board as it stood at entry, the results currently admitted to the Index and, where one exists, the estimate recorded before measurement. No text is written by hand, and no claim is made that the records do not contain. ModelCap does not run the benchmarks itself; each result links to the organisation that published it.

More launch reports