whatmodel.sh / Step 2 16K Experimental (2024-12)

Step 2 16K Experimental (2024-12)

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Step 2 16K Experimental (2024-12) is a closed-weights language model from StepFun. Licence: proprietary.

Catalogue identity notes

Exact Arena entry 'step-2-16k-exp-202412' in the October 2026 pinned dataset. Comparison identity; no public run route inferred from leaderboard inclusion.

57.5±3.3comparison only: no current run option verifiedfrom 1 benchmark result across 1 source* thinly tested, treat the rank as provisional

$ how it compares

The LLMs ranked either side of Step 2 16K Experimental (2024-12), out of 1132 with a score, then more from StepFun.

#modelscorewhereRAMreleased
347Nemotron 3 Nano 30B-A3B NVIDIA58±1api $0.087/M +local~20GB2025-12
348Ministral 3 14B Reasoning 2512 Mistral58±1local~10GB2025-12
349Step 2 16K Experimental (2024-12) StepFun58±3comparison only
350xLAM-2-3b-fc-r Salesforce58±4local~3.5GB2025-03
351MedGemma 27B Text IT Google58±4local~18GB2025-05
more from StepFun
45Step 5 Preview StepFun75±1comparison only
130Step 3.5 Flash (196B-A11B) StepFun68±1api $0.15/M +local~118GB2026-02
151Step 3.7 Flash (198B-A11B) StepFun67±1api $0.44/M +local~104GB2026-05

see all LLMs · find the ones that fit your machine

$ what it's good at

chat & writing▁▁▁▃▅▅▆█▆▆▆▂ #243 of 579 * · 95% of its score · 1 test

Beats Seed-OSS 36B Instruct, Nemotron-Cascade-14B-Thinking, Qwen3-VL 32B Instruct here, though they rank higher overall. Trails LFM2.5 1.2B Instruct, LFM2.5 1.2B Thinking, Gemma 3 12B here, though they rank lower overall.

testresultvs the others in its rundifficultycountswhy
LMArena (text)1333.6 ✎▁▁▁▂▃▃▃▆▆▇█▃ beats 169/379avg 1268▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 95%blunt

$ where to run it

No verified API or download recorded yet · add one

$ how the score was worked out · LMArena leaderboard dataset 95%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
LMArena leaderboard dataset1▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 95%CC BY 4.0
prior · pull toward the middle▇ 5%
weighted average= 57.5

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json