Step 3.5 Flash
Stored tensors can include buffers and repeated shared weights. A verified unique parameter count is unavailable. Checkpoint size is not the memory required to run the model.
GeekRadar / The comparison desk
COMPARE.
AI modelsStep 3.5 Flash
AI model Released
- Context window
- 256,000 tokens
- Maximum output
- 256,000 tokens
- Open weights
- Yes
Capability & preference
Human preference · Arena
Overall text preferences from head-to-head votes. Not factual accuracy or a task-success rate.
Recorded results
Only the same benchmark, version, workload, scenario and unit share a row. Missing results are not zero. System results include the full submitted hardware and software configuration.
Arena text preference / chinesehuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,446.33 rating
Confidence interval: 1,436.31 – 1,456.35
3667 votes
- Category
- chinese
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,435.79 rating
Confidence interval: 1,430.14 – 1,441.44
15184 votes
- Category
- coding
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,356.96 rating
Confidence interval: 1,350.27 – 1,363.66
9760 votes
- Category
- creative_writing
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,417.22 rating
Confidence interval: 1,412.49 – 1,421.95
25238 votes
- Category
- english
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,384.06 rating
Confidence interval: 1,379.30 – 1,388.82
43895 votes
- Category
- exclude_ties
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,421.53 rating
Confidence interval: 1,413.14 – 1,429.91
5336 votes
- Category
- expert
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,421.69 rating
Confidence interval: 1,407.15 – 1,436.22
2041 votes
- Category
- french
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,404.89 rating
Confidence interval: 1,386.25 – 1,423.53
1075 votes
- Category
- german
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,410.71 rating
Confidence interval: 1,406.46 – 1,414.97
37538 votes
- Category
- hard_prompts
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,420.59 rating
Confidence interval: 1,415.01 – 1,426.18
16232 votes
- Category
- hard_prompts_english
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,393.96 rating
Confidence interval: 1,387.83 – 1,400.10
11835 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,359.78 rating
Confidence interval: 1,353.57 – 1,365.98
12065 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,407.84 rating
Confidence interval: 1,398.78 – 1,416.90
4662 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,414.46 rating
Confidence interval: 1,407.94 – 1,420.99
9818 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,420.76 rating
Confidence interval: 1,409.92 – 1,431.59
3158 votes
- Category
- industry_mathematical
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,408.66 rating
Confidence interval: 1,399.14 – 1,418.18
4240 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,434.54 rating
Confidence interval: 1,429.64 – 1,439.44
22760 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,371.65 rating
Confidence interval: 1,365.94 – 1,377.35
14799 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,384.87 rating
Confidence interval: 1,379.65 – 1,390.08
19316 votes
- Category
- instruction_following
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,352.84 rating
Confidence interval: 1,328.73 – 1,376.95
679 votes
- Category
- japanese
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,350.23 rating
Confidence interval: 1,331.48 – 1,368.98
1090 votes
- Category
- korean
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,401.90 rating
Confidence interval: 1,396.78 – 1,407.02
22821 votes
- Category
- longer_query
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,408.20 rating
Confidence interval: 1,398.02 – 1,418.38
3423 votes
- Category
- math
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,404.67 rating
Confidence interval: 1,398.20 – 1,411.13
10383 votes
- Category
- multi_turn
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,385.24 rating
Confidence interval: 1,381.01 – 1,389.47
35064 votes
- Category
- non_english
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,402.75 rating
Confidence interval: 1,399.24 – 1,406.26
60305 votes
- Category
- overall
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,389.45 rating
Confidence interval: 1,373.29 – 1,405.62
1363 votes
- Category
- polish
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,384.99 rating
Confidence interval: 1,377.77 – 1,392.21
7434 votes
- Category
- russian
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-28 · rating
Step 3.5 Flash
1,418.31 rating
Confidence interval: 1,404.49 – 1,432.13
2014 votes
- Category
- spanish
- Evaluated configuration
- Step-3.5-Flash
- Evaluation type
- human preference
Arena text preference / chinesehuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,445.86 rating
Confidence interval: 1,435.84 – 1,455.89
3668 votes
- Category
- chinese
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / codinghuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,435.67 rating
Confidence interval: 1,430.02 – 1,441.32
15192 votes
- Category
- coding
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / creative_writinghuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,356.80 rating
Confidence interval: 1,350.11 – 1,363.50
9775 votes
- Category
- creative_writing
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / englishhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,417.27 rating
Confidence interval: 1,412.54 – 1,422.00
25256 votes
- Category
- english
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / exclude_tieshuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,383.87 rating
Confidence interval: 1,379.10 – 1,388.63
43923 votes
- Category
- exclude_ties
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / experthuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,421.31 rating
Confidence interval: 1,412.91 – 1,429.72
5330 votes
- Category
- expert
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / frenchhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,420.87 rating
Confidence interval: 1,406.31 – 1,435.43
2051 votes
- Category
- french
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / germanhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,405.09 rating
Confidence interval: 1,386.39 – 1,423.78
1076 votes
- Category
- german
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / hard_promptshuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,410.62 rating
Confidence interval: 1,406.36 – 1,414.87
37552 votes
- Category
- hard_prompts
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / hard_prompts_englishhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,420.70 rating
Confidence interval: 1,415.11 – 1,426.29
16241 votes
- Category
- hard_prompts_english
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_business_and_management_and_financial_operationshuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,394.02 rating
Confidence interval: 1,387.88 – 1,400.16
11840 votes
- Category
- industry_business_and_management_and_financial_operations
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_entertainment_and_sports_and_mediahuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,359.79 rating
Confidence interval: 1,353.58 – 1,366.00
12077 votes
- Category
- industry_entertainment_and_sports_and_media
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_legal_and_governmenthuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,407.80 rating
Confidence interval: 1,398.73 – 1,416.87
4662 votes
- Category
- industry_legal_and_government
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_life_and_physical_and_social_sciencehuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,414.51 rating
Confidence interval: 1,407.98 – 1,421.04
9818 votes
- Category
- industry_life_and_physical_and_social_science
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_mathematicalhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,420.79 rating
Confidence interval: 1,409.93 – 1,431.65
3155 votes
- Category
- industry_mathematical
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_medicine_and_healthcarehuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,408.58 rating
Confidence interval: 1,399.06 – 1,418.11
4241 votes
- Category
- industry_medicine_and_healthcare
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_software_and_it_serviceshuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,434.31 rating
Confidence interval: 1,429.40 – 1,439.21
22765 votes
- Category
- industry_software_and_it_services
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / industry_writing_and_literature_and_languagehuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,371.76 rating
Confidence interval: 1,366.05 – 1,377.46
14800 votes
- Category
- industry_writing_and_literature_and_language
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / instruction_followinghuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,384.72 rating
Confidence interval: 1,379.50 – 1,389.94
19325 votes
- Category
- instruction_following
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / japanesehuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,352.90 rating
Confidence interval: 1,328.71 – 1,377.10
679 votes
- Category
- japanese
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / koreanhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,350.53 rating
Confidence interval: 1,331.75 – 1,369.31
1089 votes
- Category
- korean
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / longer_queryhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,401.73 rating
Confidence interval: 1,396.61 – 1,406.85
22834 votes
- Category
- longer_query
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / mathhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,408.17 rating
Confidence interval: 1,397.99 – 1,418.36
3425 votes
- Category
- math
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / multi_turnhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,404.65 rating
Confidence interval: 1,398.18 – 1,411.13
10377 votes
- Category
- multi_turn
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / non_englishhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,385.02 rating
Confidence interval: 1,380.79 – 1,389.25
35073 votes
- Category
- non_english
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / overallhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,402.61 rating
Confidence interval: 1,399.09 – 1,406.12
60332 votes
- Category
- overall
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / polishhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,389.29 rating
Confidence interval: 1,373.13 – 1,405.45
1364 votes
- Category
- polish
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / russianhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,384.81 rating
Confidence interval: 1,377.58 – 1,392.03
7441 votes
- Category
- russian
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Arena text preference / spanishhuman preference · 2026-10-15 · rating
Step 3.5 Flash
1,418.57 rating
Confidence interval: 1,404.72 – 1,432.42
2012 votes
- Category
- spanish
- Evaluated configuration
- step-3.5-flash
- Evaluation type
- human preference
Run it locally
API cost
Standard text-token estimate. Excludes taxes, caching, batch discounts, reasoning and multimodal surcharges. Provider limits and regional availability may differ.
Step 3.5 Flash
Step 3.5 Flash / Pricing tiers
| Provider / scope | Rate | Currency | Amount / unit |
|---|---|---|---|
| SiliconFlowstepfun-ai/Step-3.5-FlashStandard | Input | CNY | 0.7 / 1M tokens |
| Output | CNY | 2.1 / 1M tokens | |
| StepFunstep-3.5-flashStandard | Cached input | CNY | 0.14 / 1M tokens |
| Input | CNY | 0.7 / 1M tokens | |
| Output | CNY | 2.1 / 1M tokens |
Capabilities
- Context window
- 256,000 tokens
- Maximum output
- 256,000 tokens
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- Yes
- Knowledge cutoff
- 2025-01
- Checkpoint tensor payload
- 371.38 GiB
Input modalities
- text
Output modalities
- text
Processing & platform
- Architecture
- step3p5
0 selected Compare 0