ai.canshu.com
中文
VRAM Calculator Advisor & TCO API Pricing Compare Models Leaderboard GPUs

Search models, GPUs, and tools

↑↓ Navigate · Enter Open · ★ Watchlist prioritized ai.canshu.com
StepFun 阶跃星辰
Open-Weight StepFun 阶跃星辰 · MoE 196.8B / A11B · Text · BF16 / FP8

Step-3.5-Flash (196B MoE) Architecture Parameters, Benchmarks & VRAM

196.8B total / 11B active MoE with 3-way Multi-Token Prediction (MTP-3) delivering up to 350 tok/s.

🧬 Step-3.5-Flash (196B MoE) Architecture & Tensor Specifications

Verified directly against HuggingFace (stepfun-ai/Step-3.5-Flash) config.json & Model Card

HF Class: Step3p5ForCausalLM

1. Backbone & Dimensions

step3p5
Total / Active Params
196.8B / 11B
Hidden Size (d_model)
4,096
FFN Intermediate Size
11,264
MoE Expert Dim
1,280
MoE Experts
288 Routed · Top-8 + 1 Shared
Activation / Norm
SiLU (SwiGLU) · ε=1e-6
Native Tensor Dtype
BF16 / FP8

2. Attention & Layer Layout

HYBRID
Num Hidden Layers
45 Layers (12 Global KV)
Hidden Layer Layout
45-Layer 3:1 SWA-to-Full Hybrid Attention · 288 Routed Experts (Top-8 Active)
Attention Heads (Q / KV)
64 Q / 8 KV (8:1 GQA)
Head Dimension
128
Linear / KV Compression
3-way Multi-Token Prediction (MTP-3) head boosts decode throughput to 100–300+ tok/s
Vision Encoder / MTP
MTP 3L

3. Context, RoPE & Tokenizer

Tokenizer
Native Context Length
262,144 tokens
RoPE Configuration
YaRN Hybrid RoPE (原生 262,144 Context)
Vocab Size
128,896 (Untied)
Special Token IDs
BOS: 0 · EOS: [1,2,128007]

📋 Step-3.5-Flash (196B MoE) Official API & System Card Specifications

Verified against official StepFun 阶跃星辰 developer documentation & pricing

API Model ID: step-3.5-flash

1. Identity & Capabilities

API Spec
API Model ID
step-3.5-flash
Release Date
2026-01-29
Weight License
Open-Weight (API Spec)
Native Reasoning
✅ Supported
Function Calling & JSON Schema
✅ 原生支持
Batch API 异步批处理
✅ Supported (50% off)

2. Context, Output & Modalities

Limits
Max Context Window
256,000 tokens (256K)
Max Output Tokens
256,000 tokens
Input Modalities
Text
Output Modalities
Text
Direct China Access
✅ Direct CN + RMB
Deployment Mode
Open-Weight + Cloud API

3. Official Pricing & Cache

CNY / 1M
Input / 1M Tokens
¥ 0.7
Cached Input / 1M
¥ 0.14 -80%
Output / 1M Tokens
¥ 2.1
Tiered Pricing Notes
StepFun: Official StepFun CN platform

🏆 Official Benchmark Results

Sourced from official System Card / Model Card & technical reports

Arena Elo ~1455 AA Index 45
🧠 Reasoning
HMMT 2025 Nov 94 %
🧠 Reasoning
IMOAnswerBench 85.4 %

International Mathematical Olympiad answer benchmark

💻 Coding
LiveCodeBench v6 86.4 %

LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code.

💻 Coding
SWE-bench Verified 74.4 %

A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.

🤖 Agentic
Terminal-Bench 2.0 51 %

Terminal-Bench 2.0 is an updated benchmark for testing AI agents' tool use ability to operate a computer via terminal.

🎛️ Official Sampling & Reasoning Parameters

Recommended generation settings & API parameter constraints

Reasoning & Agentic Coding
temperature=1 top_p=0.95 max_tokens=65,536
General Chat Domain
temperature=0.6 top_p=0.95 max_tokens=32,768

📊 Step-3.5-Flash (196B MoE) Full Benchmark Scorecard

7 public benchmarks · ranks among 154 models · synced 2026-11-20

⚔️ Compare head-to-head

Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).

⚡ Step-3.5-Flash (196B MoE) Interactive VRAM & Token Speed Calculator

Concurrency (Batch):
MoE 196.8B / A11B·45L (d=4096)·64Q/8KV·Ctx 256K
Runnable Models →
512 GB Unified·1200 GB/s·Safe Usable 460.8 GB
Smooth Fit123.02 / 460.8 GB~111.5 tok/s
3. Weight Quantization
4. Per-Request Context16K (16,384 tok)
Verified: 2026-11-20·Roofline bandwidth/compute estimate (Batch=1)
Smooth Fit
123.02/ 460.8 GB usable

Physical 512 GB · Load 26.7%

Theoretical Decode Speed
~111.5tokens/s

MoE: reads only 6.67 GB active weights/step

Max Context at Batch=1
256K/ 256K

Enable Q8/Q4 KV quantization to scale batch concurrency

Weights: 119.31 GBKV Cache (1×16K): 0.87 GBOverhead: 2.84 GB
Reserves ~10% unified memory for OS
🚀 Blazing Fast (Instant CoT)~4.5s for a 500-token response
Reference Launch CLI:
OLLAMA_CONTEXT_LENGTH=16384 ollama run step3.5:flash

Mac Studio M5 Ultra 512G Fits Step-3.5-Flash (196B MoE) · Cost & TCO

Approx. $9,499 (~$19/GB VRAM) · 190W TDP

📊 Step-3.5-Flash (196B MoE) Quantization VRAM Breakdown (16K Context)

Quantization Bits/Weight Weights KV Cache (16K) Total VRAM Quality Note
FP16 / BF16 16 bpw 393.6 GB 0.87 GB 402.8 GB Lossless Full Precision
FP8 / Q8_0 8.5 bpw 209.1 GB 0.87 GB 214.61 GB 99.9% Near-Lossless
GGUF Q6_K 6.56 bpw 161.38 GB 0.87 GB 165.93 GB Ultra High Quality
GGUF Q5_K_M 5.69 bpw 139.97 GB 0.87 GB 144.1 GB High Balanced Quality
Q4_K_M / AWQ 4.85 bpw 119.31 GB 0.87 GB 123.02 GB Sweet Spot (Recommended)
GGUF IQ4_XS 4.25 bpw 104.55 GB 0.87 GB 107.97 GB Compact 4-bit
GGUF Q3_K_M 3.91 bpw 96.19 GB 0.87 GB 99.43 GB Moderate Perplexity Loss
GGUF Q2_K 3.15 bpw 77.49 GB 0.87 GB 80.36 GB Extreme Compression

🖥️ GPU & Apple Silicon Compatibility for Step-3.5-Flash (196B MoE) (Q4_K_M)

Frequently Asked Questions

What are the exact architecture parameters of Step-3.5-Flash (196B MoE)?

Step-3.5-Flash (196B MoE) uses Step3p5ForCausalLM with 45 layers (45-Layer 3:1 SWA-to-Full Hybrid Attention · 288 Routed Experts (Top-8 Active)), hidden_size=4096, 64 Q heads / 8 KV heads (head_dim=128), and vocab_size=128,896.

How much VRAM does Step-3.5-Flash (196B MoE) require?

At 4-bit quantization (Q4_K_M, 16K context), Step-3.5-Flash (196B MoE) requires 123.02 GB VRAM. Full FP16 requires 402.8 GB.