FITMYLLM · AUGUST 17, 2026 · OPEN METHODOLOGY · COMMUNITY BENCHMARKS
LIVE · UPDATED DAILY
201.37B
Parameters (11B active)
Quantization Options Select your GPU for speed estimates Context length: 4K 8K 16K 32K 64K 128K 256K
Quant Bits VRAM @ 16K Quality IQ2_XXS 2.38 61.0 GB
60.4 + 0.6 KV
low IQ2_M 2.93 74.8 GB
74.2 + 0.6 KV
low Q2_K 3.16 80.6 GB
80.0 + 0.6 KV
low IQ3_XXS 3.25 82.9 GB
82.3 + 0.6 KV
low IQ3_XS 3.5 89.2 GB
88.6 + 0.6 KV
low Q3_K_S 3.64 92.7 GB
92.1 + 0.6 KV
low IQ3_M 3.76 95.7 GB
95.1 + 0.6 KV
low Q3_K_M 4 101.8 GB
101.2 + 0.6 KV
low Q3_K_L 4.3 109.3 GB
108.7 + 0.6 KV
moderate IQ4_XS 4.46 113.3 GB
112.8 + 0.6 KV
moderate Q4_K_S 4.67 118.6 GB
118.0 + 0.6 KV
moderate Q4_K_M 4.89 124.2 GB
123.6 + 0.6 KV
good Q5_K_S 5.57 141.3 GB
140.7 + 0.6 KV
good Q5_K_M 5.7 144.5 GB
144.0 + 0.6 KV
good Q6_K 6.56 166.2 GB
165.6 + 0.6 KV
excellent Q8_0 8.5 215.0 GB
214.4 + 0.6 KV
lossless FP16 16 403.8 GB
403.2 + 0.6 KV
lossless
Select your GPU above to see speed estimates and compatibility for each quantization.
Too big for a single GPU — plan a multi-GPU deployment
Even the lightest quant needs ~61 GB. Size GPUs, replicas, TCO and scaling for a production setup. Open in Enterprise →
▸ READY TO RUN THIS? RENT BY THE HOUR
RENT A GPU AND RUN STEP 3.7 FLASH NOW
Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.
Community Ratings Chat Coding Reasoning Creative Vision Roleplay Agentic
Loading ratings...
Run this model IQ2_XXS — 60.4 GB VRAM IQ2_M — 74.2 GB VRAM Q2_K — 80.0 GB VRAM IQ3_XXS — 82.3 GB VRAM IQ3_XS — 88.6 GB VRAM Q3_K_S — 92.1 GB VRAM IQ3_M — 95.1 GB VRAM Q3_K_M — 101.2 GB VRAM Q3_K_L — 108.7 GB VRAM IQ4_XS — 112.8 GB VRAM Q4_K_S — 118.0 GB VRAM Q4_K_M — 123.6 GB VRAM Q5_K_S — 140.7 GB VRAM Q5_K_M — 144.0 GB VRAM Q6_K — 165.6 GB VRAM Q8_0 — 214.4 GB VRAM FP16 — 403.2 GB VRAM
Ollama llama.cpp vLLM LM Studio KoboldCpp Jan Docker
▸ Easiest way to get started · Beginners
DOCS ↗ curl -fsSL https://ollama.com/install.sh | shCOPY
$ ollama run other:201b-q4_K_MCOPY
Tag may need adjustment — check ollama.com/library/other for available tags.
▸ SETUP GUIDE >_
Auto-setup with fitmyllm CLI Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.
Auto-detect GPU Live tok/s in chat Speed benchmarks 9 inference engines
GPUs that can run this model At Q4_K_M quantization. Sorted by minimum VRAM.
Find the best GPU for Step 3.7 Flash
Build Hardware for Step 3.7 Flash ▸ COLOPHON FITMYLLM · INDEPENDENT · DATA-DRIVEN
FITMYLLM · EST. 2025 · © 2026
RECOMMENDATIONS FROM PUBLISHED MATH, CORRECTED BY THE COMMUNITY — 30.
▸ SPEC SHEET
Step 3.7 Flash — 201.37B MoE. ▸ SPECIFICATIONS
PARAMETERS 201.37B (11B active)
ARCHITECTURE Mixture of Experts
CONTEXT LENGTH 256K tokens
CAPABILITIES chat, coding, reasoning, multilingual, vision, math
RELEASE DATE 2026-05-29
PROVIDER StepFun
FAMILY other ▸ VRAM REQUIREMENTS
QUANT BPW VRAM QUALITY IQ2_XXS 2.38 60.4 GB 65% IQ2_M 2.93 74.2 GB 75% Q2_K 3.16 80.0 GB 78% IQ3_XXS 3.25 82.3 GB 82% IQ3_XS 3.5 88.6 GB 84% Q3_K_S 3.64 92.1 GB 85% IQ3_M 3.76 95.1 GB 86% Q3_K_M 4 101.2 GB 88% Q3_K_L 4.3 108.7 GB 90% IQ4_XS 4.46 112.8 GB 92% Q4_K_S 4.67 118.0 GB 93% Q4_K_M 4.89 123.6 GB 94% Q5_K_S 5.57 140.7 GB 96% Q5_K_M 5.7 144.0 GB 96% Q6_K 6.56 165.6 GB 97% Q8_0 8.5 214.4 GB 100% FP16 16 403.2 GB 100%
§ 01 BENCHMARK SCORES
GPQA Diamond 80.9
HLE 21.4
AA Intelligence 30.9
AA Coding 39.6
aa_ifbench 67.3
aa_terminal_bench 35.6
aa_tau2 98.5
aa_scicode 40.0
aa_lcr 69.7
§ 03 COMPATIBLE GPUs
29 @ Q4_K_M Feedback