DeepSeek V4 Pro 0813 (Non-reasoning) vs Granite 4.2 30B
Which is better for the money: DeepSeek V4 Pro 0813 (Non-reasoning) by DeepSeek or Granite 4.2 30B by IBM? 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
DeepSeek V4 Pro 0813 (Non-reasoning) scores 7.7 points higher (20.4 vs 12.7) but costs 4.1× as much to run the benchmark ($451 vs $111; both mid-range class). They are not alternatives at one budget: pick DeepSeek V4 Pro 0813 (Non-reasoning) when you need intelligence above 12.7, and Granite 4.2 30B when 12.7 is enough.
| Metric | DeepSeek V4 Pro 0813 (Non-reasoning) | Granite 4.2 30B | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 20.4 | 12.7 | −37.7% Granite 4.2 30B is 37.7% less intelligent |
| Total AA cost (lower is better) | $450.76 | $110.83 | −75.4% Granite 4.2 30B is 75.4% cheaper |
| End-to-end time (lower is better) | 6.74s | 55.20s | +719.0% Granite 4.2 30B is 719.0% slower |
| Price per 1M tokens (in / out) | $1.32 / $3.96 | $0.16 / $0.65 | |
| Input modalities | Text | Text |
Granite 4.2 30B vs DeepSeek V4 Pro 0813 (Non-reasoning): 37.7% less intelligence, 75.4% less cost to run the AA benchmark, and 719.0% slower end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek V4 Pro 0813 (Non-reasoning) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Medium) scores 47.80 for $361.
- Granite 4.2 30B is off the frontier: at the same or lower cost, MiMo-V2.6-Flash scores 37.90 for $110.
DeepSeek V4 Pro 0813 (Non-reasoning)Granite 4.2 30BToday'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-non-reasoning/granite-4-2-30b · Markdown: /api/v1/compare/deepseek-v4-pro-non-reasoning/granite-4-2-30b?format=md · llms.txt · agent skill