โ† Back to Model Hub/SIDE-BY-SIDE REVIEW
SHARE THIS:

Kimi K2 0905 vs Step 3.7 Flash

How do these models stack up? Below is an expert side-by-side comparison of specifications, context window capacity, live pricing per million tokens, and standardized benchmark scores for Kimi K2 0905 and Step 3.7 Flash.

Moonshot AI

Kimi K2 0905

Kimi K2 0905 is the September update of [Kimi K2 0711](moonshotai/kimi-k2). It is a large-scale Mixture-of-Experts (MoE) language model developed by Moonshot AI, featuring 1 trillion total parameters with 32...

View Full Specs
StepFun

Step 3.7 Flash

Step 3.7 Flash is StepFun's latest high-efficiency multimodal Mixture-of-Experts model. It pairs a 196B-parameter language backbone with a vision encoder for native image and video understanding, activating roughly 11B parameters...

View Full Specs

Technical Specifications

SpecificationKimi K2 0905Step 3.7 Flash
ProviderMoonshot AIStepFun
Context Window262,144 tokens256,000 tokens
Agent SuitabilityN/AN/A
Time to First Token (TTFT)N/AN/A
Deployment Modelmanaged apimanaged api
Production Stabilitystablestable
API AvailableYesYes
Released Date2025-09-042026-05-28

API Pricing Comparison

Input Price per Million Tokens

Kimi K2 0905

$0.60

Step 3.7 Flash

$0.20

Output Price per Million Tokens

Kimi K2 0905

$2.50

Step 3.7 Flash

$1.15

Want to test both models live?

Run side-by-side prompt prompts in our dynamic Sandbox. Check execution speeds, latency metrics, and compute actual costs in real-time.

Benchmark Performance Metrics

Scores show the raw performance percentages verified across key evaluation suites. Higher bars indicate superior accuracy and capability in that domain.

MMLUGeneral knowledge & multi-task understanding
N/Avs8450.0%
Kimi K2 0905
Step 3.7 Flash
HumanEvalPython coding & logic synthesis
N/Avs8600.0%
Kimi K2 0905
Step 3.7 Flash
MATHComplex mathematical problem solving
N/Avs6680.0%
Kimi K2 0905
Step 3.7 Flash
GPQAGraduate-level expert reasoning
N/Avs4400.0%
Kimi K2 0905
Step 3.7 Flash
HellaSwagCommonsense reasoning and inference
N/Avs8580.0%
Kimi K2 0905
Step 3.7 Flash
MT-BenchMulti-turn conversation flow quality
N/Avs890.0%
Kimi K2 0905
Step 3.7 Flash

Kimi K2 0905 Quirks & Gotchas

No developer gotchas reported.

Step 3.7 Flash Quirks & Gotchas

No developer gotchas reported.