DeepSeek V4 Pro 0813 (Max) vs Step 5 Preview
Which is better for the money: DeepSeek V4 Pro 0813 (Max) by DeepSeek or Step 5 Preview by StepFun? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-25.
Trade-off
Verdict
Depends on your budget
Step 5 Preview scores 7.7 points higher (43.7 vs 36.0) but costs 1.1× as much to run the benchmark ($1190 vs $1122; both premium class). Pick Step 5 Preview when the extra intelligence is worth the premium, otherwise DeepSeek V4 Pro 0813 (Max).
| Metric | DeepSeek V4 Pro 0813 (Max) | Step 5 Preview | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 36.0 | 43.7 | +21.4% Step 5 Preview is 21.4% more intelligent |
| Total AA cost (lower is better) | $1,122.27 | $1,190.26 | +6.1% Step 5 Preview is 6.1% more expensive |
| End-to-end time (lower is better) | 25.87s | 12.01s | −53.6% Step 5 Preview is 53.6% faster |
| Price per 1M tokens (in / out) | $1.32 / $3.96 | $1.00 / $2.70 | |
| Input modalities | Text | Text, Image |
Step 5 Preview vs DeepSeek V4 Pro 0813 (Max): 21.4% more intelligence, 6.1% more cost to run the AA benchmark, and 53.6% faster end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek V4 Pro 0813 (Max) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
- Step 5 Preview is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
DeepSeek V4 Pro 0813 (Max)Step 5 PreviewToday's Pareto frontierOpen-weights frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/deepseek-v4-pro/step-5 · Markdown: /api/v1/compare/deepseek-v4-pro/step-5?format=md · llms.txt · agent skill