Inference
EN
StepFun

Step 3.7 Flash

Step 3.7 Flash is a reasoning model from StepFun. 21 benchmarks count toward its index, in 6 categories.

  • Open weights
  • Reasoning
  • Input: text · image · video
  • Released 2026-05
RankedEnough results for an index.
IndexAll categories, out of 100.55.3±5.0Median 51.0
CoverageIndex weight that has results.85%
SpeedOutput tokens per second.112/sMedian 67/s
First tokenSeconds to the first output token.3.5 sMedian 1.3 s
Input / 1MUS dollars per 1M input tokens.$0.20Median $0.60
Output / 1MUS dollars per 1M output tokens.$1.15Median $2.55
ContextMaximum tokens in one request.262KMedian 262K
EloLMArena rating and rank.—Median 1422

Details

From OpenRouter
API IDThe model ID on OpenRouter.
stepfun/step-3.7-flash
Max outputMaximum output tokens per request.
230K
Knowledge cutoffLast date of training data.
—
ToolsTool calls on OpenRouter.
Yes
JSON outputOutput follows a given JSON schema.
Yes
Cache read / 1MUS dollars per 1M cached tokens.
$0.04
WeightsLink to the published weights.
Hugging Face (opens in a new tab) ↗
RetiresDate OpenRouter removes the model.
—

Where to run

APIs that serve this model.
4 providers
ProviderTypeMaker, cloud, host or router.InputOutputContext
Deep InfraDeprecatedHost$0.20$1.15262K
Novita AIHost$0.20$1.15262K
OpenRouterRouter$0.20$1.15262K
StepFunMaker$0.20$1.15256K

US dollars per 1M tokens, list price. From models.dev (opens in a new tab) (MIT).

Run locally

201.37B parameters
16-bit
423.9 GB
8-bit
225.7 GB
4-bit
129.2 GB
Weights
Hugging Face (opens in a new tab) ↗

Memory need with a short context. The 4-bit copy needs more than 128 GB. Compare local models

Capabilities

Index per category, out of 100.
Out of 100
AgenticMulti-step tasks with tools.10 counted59.672 / 218
CodingCode writing and repair.2 counted54.9100 / 209
ReasoningLogic problems and puzzles.2 counted49.6119 / 227
MultimodalTasks with images and text.3 counted59.342 / 84
KnowledgeFacts and expert knowledge.3 counted46.8159 / 268
MultilingualTasks in many languages.None counted——
InstructionTasks with strict prompt rules.1 counted51.9—
MathMath problems.None counted——

Results

21 counted
BenchmarkThe benchmark name.ResultThe score from the source.PlacePlace among results on this benchmark.
τ²-Bench Tool-Agent-User EvaluationAgentic98.5%3 / 149
SimpleVQAMultimodal79.2%2 / 18
DeepSearchQAAgentic92.8%5 / 24
V*Multimodal95.3%4 / 14
Claw-EvalAgentic67.1%15 / 42
Gert Labs Composite Game BenchmarkAgentic51.6%19 / 52
Artificial Analysis IFBenchInstruction67.3%54 / 146
Artificial Analysis MMMU-ProMultimodal75.3%52 / 113
Artificial Analysis Long Context ReasoningReasoning73.7%100 / 217
Critical Physics TasksReasoning2.3%104 / 214
Artificial Analysis GPQA DiamondKnowledge80.9%97 / 199
Artificial Analysis Humanity's Last ExamKnowledge21.4%110 / 222
ToolathlonAgentic49.5%11 / 22
GDPval-AA normalizedAgentic25.8%69 / 133
BrowseCompAgentic75.8%37 / 63
SWE-bench ProCoding56.3%49 / 83
Artificial Analysis Omniscience AccuracyKnowledge25.8%101 / 164
Artificial Analysis SciCodeCoding43.9%78 / 114
Terminal-Bench 2.1 · BenchLMAgentic59.5%65 / 86
APEX-Agents-AAAgentic14.8%23 / 30
Humanity's Last Exam with toolsAgentic47.2%30 / 36

Sources

BenchLMCC BY-NC 4.0OpenRouterTerms restrict reusemodels.devMIT

History

5 releases
Index56.2 → 55.3
23 Oct 202612 Nov 2026
Output / 1M$1.15 → $1.15
23 Oct 202612 Nov 2026

Questions

How much does Step 3.7 Flash cost?
$0.20 per 1M input tokens, $1.15 per 1M output tokens.
How good is Step 3.7 Flash?
Its index is 55.3 out of 100. That is place 99 of 283 ranked models.
What is Step 3.7 Flash best at?
Agentic: 59.6 out of 100, place 72 of 218.
How fast is Step 3.7 Flash?
112 output tokens per second, 3.5 s to the first token.
What is the context window of Step 3.7 Flash?
262K tokens.
Is Step 3.7 Flash open source?
Yes. Its weights are published.
When was Step 3.7 Flash released?
May 2026.

Badge

Step 3.7 Flash on Inference360: place 99 of 283, index 55.3
HTML
<a href="https://inference360.ai/models/step-3-7-flash"><img src="https://inference360.ai/badges/step-3-7-flash.svg" alt="Step 3.7 Flash on Inference360: place 99 of 283, index 55.3" height="20"></a>
Markdown
[![Step 3.7 Flash on Inference360: place 99 of 283, index 55.3](https://inference360.ai/badges/step-3-7-flash.svg)](https://inference360.ai/models/step-3-7-flash)

Compare with

GPT-5.155.3Qwen3.5 397B (Reasoning)55.3Inkling-Small55.1dots3-note Preview63.4Step 5 Preview68.4

Lower price

dots3-note Preview63.4 · FreeApodex 1.1 Mini55.7 · FreeGLM-5.3-Flash63.1 · $0.078/1M

More from StepFun

All models
Step 5 Preview68.4Step 3.5 FlashUnrankedStep-AudioUnrankedStep-Audio-Chat 130BUnranked