Step 3.5 Flash
Feb 2026
Older versionReplace withStep 3.7 FlashNewer version
Nemotron 3.5 Super VLCheaper, as good
GPT-5.6 LunaBetter
MiMo v2.6 ProOpen weights
Step 3.7 FlashNewer versionNemotron 3.5 Super VLCheaper, as goodGPT-5.6 LunaBetterMiMo v2.6 ProOpen weightsStep 3.5 FlashCompare inon
Benchmarks of Step 3.5 Flash11
BenchmarkByScoreRank
Overall1
Knowledge4
Science2
Coding1
Documents1
Legacy · saturated or no longer reported2
No benchmark matches
Rankingin thinking mode?
Retired models are left out of the current rankings.
Alternatives?
ModelScoreCostPer taskWhat changes
CheaperNemotron 3.5 Super VLthinkingOpen+59%-12%$0.19BetterGPT-6 Lunalow+30%+33%$0.005Function calling only with none · Chat Completions+3OpenMiMo v2.6 Prothinking×2.8+262%$0.13Pass back reasoning_content+3
CheaperLing 3.0 Flash VLthinkingOpen+48%-40%$0.06 / $0.18
CheaperSolar Mini 4+45%-42%$0.51ModelScoreCostPer taskWhat changes
Nemotron 3.5 Super VLthinkingOpen+59%-12%$0.19Ling 3.0 Flash VLthinkingOpen+48%-40%$0.06 / $0.18Solar Mini 4+45%-42%$0.51Ling 3.0 Flash FinthinkingOpen+36%-40%$0.06 / $0.18Ling 3.0 FlashthinkingOpen+21%-40%$0.028ModelScoreCostPer taskWhat changes
GPT-6 Lunalow+30%+33%$0.005Function calling only with none · Chat Completions+3GPT-5.6 Lunalow+26%+200%$0.01
Claude Haiku 5.5low+77%+33%$0.024Remove sampling parameters+5Ling 3.0 FlashthinkingOpen+21%-40%$0.028MiMo v2.5thinkingOpen+31%+17%$0.042ModelScoreCostPer taskWhat changes
MiMo v2.6 ProthinkingOpen×2.8+262%$0.13Pass back reasoning_content+3
DeepSeek V4.1 Flash 1017maxOpen×2.7+250%$0.26Pass back reasoning_content · Chat Completions+4
GLM 5.3maxOpen×2.7+1,333%$2.01Reasoning cannot be disabled+3Step 5 PreviewthinkingOpen×2.6+850%$1.03web_search not available · Chat Completions+2
Kimi K3maxOpen×2.6+3,900%$2.00Sampling parameters are fixed+3Run it yourself
Fits onhardware needs not verified yet
Licencenot verified
History
4 versionsBenchmarks
all categoriesChat & Knowledge
AA Omniscience Index-41.8
GPQA Diamond83.1
Coding
Agents & Tool Use
τ²-bench · AA94.4
Writing & Language
FAQ
How much does Step 3.5 Flash cost?›
$0.10 per 1M input tokens and $0.30 per 1M output tokens; cached input $0.02 per 1M.
Is Step 3.5 Flash free to use?›
There is no free API; the cheapest paid price is $0.10 per 1M input tokens. The weights are open, so you can run it yourself.
How can I use Step 3.5 Flash?›
Through the StepFun API.
What is the context window of Step 3.5 Flash?›
256K tokens.
When was Step 3.5 Flash released? Is there a newer version?›
Released 2026-02-02. It has been superseded by Step 3.5 Flash.
Is Step 3.5 Flash open source?›
Yes, the weights are open.
Can I run Step 3.5 Flash locally?›
The weights are open and can be self-hosted; hardware needs are not confirmed yet.
What is Step 3.5 Flash?›
Step 3.5 Flash is an open-weights language model from StepFun, released Feb 2026.
What are the best alternatives to Step 3.5 Flash?›
Nemotron 3.5 Super VL (cheaper), GPT-6 Luna (better), MiMo v2.6 Pro (open weights).