ModelsCompareBest forBenchmarksStatusPricingAPI

Step 3.5 Flash

StepFun·Pending·Released 2026-02-12text
Context256k
Max output-
Input / 1M-
Output / 1M-
Cached in-
Benchmarks18

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

AIME 2025 (MathArena)#2/2398.3% ↗
AIME 2026 (MathArena)#6/2196.7% ↗
LMArena Elo#43/621403.7 ↗
AIME 2025#1/1597.3% ↗
BrowseComp#12/1551.6% ↗
BrowseComp (with context manager)69% ↗
BrowseComp-ZH#1/266.9% ↗
BrowseComp-ZH (with context manager)73.7% ↗
GAIA (no file)84.5% ↗
HMMT 2025 (Feb/Nov average)96.2% ↗
IMOAnswerBench#2/285.4% ↗
LiveCodeBench v6#4/1186.4% ↗
ResearchRubrics#3/365.3% ↗
SWE-bench Verified#19/4574.4% ↗
Tau2-Bench#1/288.2% ↗
Terminal-Bench 2.0#8/1351% ↗
xbench-DeepSearch 2025.0583.7% ↗
xbench-DeepSearch 2025.1056.3% ↗

Compare

vs GPT-6 Astravs GPT-5.6 Solvs GPT-5.6 Terravs GPT-5.6 Lunavs Claude Fable 5.1vs Claude Opus 5vs Claude Sonnet 5vs Claude Haiku 4.5

More from StepFun

Step 3.7 Flash

Access Step 3.5 Flash and every other model through one endpoint with automatic failover: Respan gateway.