Skip to content
Model Pareto

Step 3.7 Flash

StepFunOpen weightsartificialanalysis.ai ↗
Compare with nearest rivalsCompare with…
Input
$0.20/M tok
Output
$1.15/M tok
Blended
$0.438/M tok · 3:1
Speed
190tok/s
Context
262Ktokens

Long-tail import from Artificial Analysis (2026-09-24); AA variant: Step 3.7 Flash. Price = AA median host price.

Rankings

Self-hostingEstimated

How we estimate →

What it would cost to run these open weights yourself on rented GPUs, compared with the API price above.

Hardware
2× B200 180GB
FP8 weights
Throughput
~8,082 tok/s
many requests batched
Price
$0.15–1.38 /M tok
busy → light use
On your own machine
Mac Studio M5 Max 128GB (int4, ~78 tok/s single-stream)
single consumer GPU or Mac
Assumptions (5)
  • FP8 weights (198B params, 11B active per token) + 25% KV-cache headroom ≈ 248 GB
  • 2× B200 180GB at $5.98–$14.24/GPU-hour on-demand (2026-09-24)
  • ~8,082 output tok/s aggregate at batch 231 (bandwidth-bound); MoE compute scales with active params
  • Blended 3:1 input:output; prefill ~193,295 tok/s
  • Low = 75% utilization at the low GPU price; high = 20% at the high price

Benchmark scores

~ italic, dashed = no published score yet, estimated from related benchmarks.

Coding

SWE-bench VerifiedEstimated
~#10 of 14 (estimated)~75.0%
  1. #1Claude Opus 596.0%
  2. #2Claude Sonnet 585.2%
  3. #9Xing4.0-29B-A4B75.0%
  4. ~#10Step 3.7 FlashEstimated75.0%
  5. #10Claude Haiku 4.573.3%
No published score — estimated from related benchmarks
SWE-bench Pro (Public)Estimated
~#14 of 19 (estimated)~54.3%
  1. #1Claude Opus 5.589.9%
  2. #2Claude Sonnet 5.581.3%
  3. #13Hy357.9%
  4. ~#14Step 3.7 FlashEstimated54.3%
  5. #14Inkling54.3%
No published score — estimated from related benchmarks
DeepSWE v1.1Estimated
~#28 of 31 (estimated)~49.5%
  1. #1DeepSeek V4.1 Flash74.2%
  2. #2Grok 4.772.6%
  3. #27Qwen3.8 Max (0902)51.0%
  4. ~#28Step 3.7 FlashEstimated49.5%
  5. #28Qwen3.8 27B42.2%
No published score — estimated from related benchmarks
Terminal-Bench 4.0Estimated
~#26 of 84 (estimated)~11.4%
  1. #1Claude Sonnet 5.563.6%
  2. #2Claude Opus 5.559.6%
  3. #23Kimi K312.6%
  4. ~#26Step 3.7 FlashEstimated11.4%
  5. #26Qwen3.8 2.4T A95B11.1%
No published score — estimated from related benchmarks
LiveCodeBenchEstimated
~#8 of 20 (estimated)~76.1%
  1. #1Qwen3.8-Omni-Flash92.6%
  2. #2Qwen3.8 Flash-Next91.9%
  3. #7gpt-oss-20b77.7%
  4. ~#8Step 3.7 FlashEstimated76.1%
  5. #8Mistral Small 464.0%
No published score — estimated from related benchmarks
AA Coding Agent IndexEstimated
~last of 15 (estimated)below measured range (<41.1)
  1. #1Claude Opus 5.566.0
  2. #2Claude Fable 5.162.2
  3. #14GPT-6 Luna41.1
  4. ~#15Step 3.7 FlashEstimated<41.1
No published score — estimated from related benchmarks

Writing & chat

MMMLUEstimated
~#3 of 5 (estimated)~87.5%
  1. #1Gemini 3.1 Pro92.6%
  2. #2Gemma 4 31B88.4%
  3. ~#3Step 3.7 FlashEstimated87.5%
  4. #3Claude Haiku 4.583.0%
No published score — estimated from related benchmarks
LMArena Text, Non-English (Elo)Estimated
~#33 of 54 (estimated)~1416 Elo
  1. #1Claude Opus 5.51501
  2. #2Claude Fable 5.1Few votes1496
  3. #32Muse GlimmerFew votes1418
  4. ~#33Step 3.7 FlashEstimated1416
  5. #33MiMo-V2.51411
No published score — estimated from related benchmarks
LMArena Text (Elo)Estimated
~#33 of 54 (estimated)~1430 Elo
  1. #1Claude Opus 5.51509
  2. #2Claude Fable 5.11501
  3. #32MiMo-V2.51434
  4. ~#33Step 3.7 FlashEstimated1430
  5. #33Mistral Medium 3.51426
No published score — estimated from related benchmarks
EQ-Bench Creative Writing v3 (Elo)Estimated
~#13 of 26 (estimated)~1815 Elo
  1. #1GPT-6 Astra2173
  2. #2Claude Fable 5.12162
  3. #12Qwen3.8 2.4T A95B1843
  4. ~#13Step 3.7 FlashEstimated1815
  5. #13Muse Glimmer1798
No published score — estimated from related benchmarks

Vision

MMMUEstimated
~#2 of 3 (estimated)~71.3%
  1. #1Claude Haiku 4.573.2%
  2. ~#2Step 3.7 FlashEstimated71.3%
  3. #2Llama 4 Maverick69.4%
No published score — estimated from related benchmarks
CharXiv ReasoningEstimated
~#6 of 9 (estimated)~78.6%
  1. #1Gemini 3.8 Flash86.2%
  2. #2Kimi K384.8%
  3. #5Muse Glimmer78.8%
  4. ~#6Step 3.7 FlashEstimated78.6%
  5. #6Inkling78.1%
No published score — estimated from related benchmarks
OmniDocBench v1.5Estimated
~#5 of 8 (estimated)~84.9%
  1. #1MiniMax-M391.6%
  2. #2Kimi K391.1%
  3. #4Gemini 3.1 Pro85.3%
  4. ~#5Step 3.7 FlashEstimated84.9%
  5. #5Claude Haiku 4.579.6%
No published score — estimated from related benchmarks
GDP.pdf (AA)Estimated
~#23 of 43 (estimated)~13.3%
  1. #1GPT-6 Astra31.0%
  2. #2Muse Spark 1.326.6%
  3. #22GPT-5.4 mini13.4%
  4. ~#23Step 3.7 FlashEstimated13.3%
  5. #23Claude Sonnet 513.2%
No published score — estimated from related benchmarks
ChartographyEstimated
~#14 of 23 (estimated)~21.2%
  1. #1GPT-6 Astra71.0%
  2. #2Claude Opus 5.566.3%
  3. #13Gemini 3.1 Pro26.1%
  4. ~#14Step 3.7 FlashEstimated21.2%
  5. #14Grok 4.620.5%
No published score — estimated from related benchmarks

Hard reasoning

AIME (latest)Estimated
~#5 of 25 (estimated)~90.7%
  1. #1Inkling97.1%
  2. #2Muse Glimmer94.7%
  3. #4gpt-oss-120b93.4%
  4. ~#5Step 3.7 FlashEstimated90.7%
  5. #5Xing4.0-29B-A4B90.0%
No published score — estimated from related benchmarks
AA Intelligence IndexEstimated
~#37 of 75 (estimated)~24.7
  1. #1Claude Opus 5.557.6
  2. #2Claude Sonnet 5.556.0
  3. #36Inkling25.0
  4. ~#37Step 3.7 FlashEstimated24.7
  5. #37Ling-3.0-flash-VL24.6
No published score — estimated from related benchmarks

Agents

OSWorld-VerifiedEstimated
~#10 of 12 (estimated)~66.3%
  1. #1Kimi K384.8%
  2. #2Qwen3.8 27B84.3%
  3. #9Nex-N2.5-Mini71.2%
  4. ~#10Step 3.7 FlashEstimated66.3%
  5. #10Muse Glimmer65.9%
No published score — estimated from related benchmarks
OSWorld 2.0Estimated
~last of 5 (estimated)below measured range (<47.5%)
  1. #1GLM-5.3 Flash59.1%
  2. #2Kimi K358.3%
  3. #4GPT-5.547.5%
  4. ~#5Step 3.7 FlashEstimated<47.5%
No published score — estimated from related benchmarks
BrowseCompEstimated
~#15 of 18 (estimated)~64.9%
  1. #1Atria Dawn Preview92.5%
  2. #2GPT-6 Astra91.5%
  3. #14Inkling77.1%
  4. ~#15Step 3.7 FlashEstimated64.9%
  5. #15Mistral Medium 3.548.6%
No published score — estimated from related benchmarks
AutomationBench-AAEstimated
~#40 of 80 (estimated)~15.9%
  1. #1Claude Opus 5.569.5%
  2. #2DeepSeek V4.1 Flash68.9%
  3. #39Qwen3.7 Plus17.4%
  4. ~#40Step 3.7 FlashEstimated15.9%
  5. #40Ling-3.0-flash-VL15.7%
No published score — estimated from related benchmarks