The Model Index
Up to four models

DeepSeek-V4.1-Flash vs Step 5 Preview

On the index (GPT-4 = 100), DeepSeek-V4.1-Flash scores 163 and Step 5 Preview 163. Side by side on the same third-party tests, with price and context. The link in your address bar shares this exact comparison.

Next to DeepSeek-V4.1-Flash
Presets
DeepSeek-V4.1-Flash
163147–179
WeightsOpen
ReleasedSep 2026
Price in / out$0.051 / $0.57
Context1.05M
Core tests taken6 of 9
Step 5 Preview
163145–180
WeightsClosed
ReleasedSep 2026
Price in / out$1.0 / $2.7
Context1M
Core tests taken2 of 9

DeepSeek-V4.1-Flash and Step 5 Preview are too close to call (1 point apart, within the uncertainty), ahead on 1 of 2 shared core tests, and costs 7.9× less.

Test by testthird-party results

DeepSeek-V4.1-FlashStep 5 Preview
0%25%50%75%100%MATHSProofBench v1.1ProofBench v1.1CODINGTerminal-Bench 4.0Terminal-Bench 4.0Code MigrationCode MigrationIOIIOIVibe Code Bench v1.1Vibe Code Bench v1.1AGENTSBioMysteryBenchBioMysteryBenchSRE BenchSRE BenchTerminal-Bench-ScienceTerminal-Bench-SciencePROFESSIONAL WORKFinance Agent (v2)Finance Agent (v2)EMBEMBLegal Research BenchLegal Research BenchPublic Benefits Bench v1.1Public Benefits Bench v1.1Tax Agent BenchTax Agent BenchNOVEL PROBLEMSMysteryMechanismMysteryMechanism
Every result as a table
TestDeepSeek-V4.1-FlashStep 5 Preview
Maths
ProofBench v1.154.0%42.0%
Coding
Terminal-Bench 4.019.7%31.8%
Code Migration45.6%39.0%
IOI40.3%38.1%
Vibe Code Bench v1.184.7%70.4%
Agents
BioMysteryBench67.8%67.4%
SRE Bench0.8%0.4%
Terminal-Bench-Science0.0%2.9%
Professional work
Finance Agent (v2)53.5%50.7%
EMB57.2%57.1%
Legal Research Bench41.3%40.4%
Public Benefits Bench v1.164.3%58.1%
Tax Agent Bench26.6%27.3%
Novel problems
MysteryMechanism21.2%18.5%

Tests at least two of these models have taken. Bold rows are the core tests.

This page as a chartDownload PNG