StepFun: Step 3.5 Flash
Stepfun
Step 3.5 Flash is StepFun's most capable open-source foundation model. Built on a sparse Mixture of Experts (MoE) architecture, it selectively activates only 11B of its 196B parameters per token....
Pricing
Source-reported prices are shown in their native units. No missing fee is estimated.
Input
$0.100
USD / 1M input tokens
Output
$0.300
USD / 1M output tokens
Capabilities
A check means the source explicitly lists support; a dash means the source did not publish a value.
Text input Supported
Text output Supported
Vision Not listed
Tools Supported
Reasoning Supported
Structured outputs Not listed
Audio Not listed
Image generation Not listed
Embeddings Not listed
Alternatives
Suggestions use deterministic rules: overlapping modalities, a context window within 2×, and shared source-reported signals. No AI-generated recommendations.
Similar models
Amazon: Nova Lite 1.0
amazon/nova-lite-v1
Input
$0.060
Intelligence
6.7
Similar category, context, price or shared benchmark signals.
Google: Gemma 3 27B
google/gemma-3-27b-it
Input
$0.080
Intelligence
4.9
Similar category, context, price or shared benchmark signals.
Inception: Mercury 2.5
inception/mercury-2.5
Input
$0.040
Intelligence
12.3
Similar category, context, price or shared benchmark signals.
Meta: Llama 3.1 70B Instruct
meta-llama/llama-3.1-70b-instruct
Input
$0.400
Intelligence
6.6
Similar category, context, price or shared benchmark signals.
Higher-performance alternatives
Higher on at least one shared benchmark and never lower on the other shared benchmarks, with similar context.
SpaceXAI: Grok Build 0.1
x-ai/grok-build-0.1
Input
$1.00
Intelligence
27.2
Higher source-reported performance on shared benchmark dimensions.
Tencent: Hy3 preview
tencent/hy3-preview
Input
$0.180
Intelligence
25.3
Higher source-reported performance on shared benchmark dimensions.
OpenAI: GPT-5.2-Codex
openai/gpt-5.2-codex
Input
$1.75
Intelligence
28.5
Higher source-reported performance on shared benchmark dimensions.
Qwen: Qwen3.5-27B
qwen/qwen3.5-27b
Input
$0.270
Intelligence
22.9
Higher source-reported performance on shared benchmark dimensions.