DeepSeek V3 (Dec '24) vs Mercury 2
Which is better for the money: DeepSeek V3 (Dec '24) by DeepSeek or Mercury 2 by Inception? 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
Mercury 2 scores 2.3 points higher (10.8 vs 8.5) but costs 4.1× as much to run the benchmark ($105 vs $25.4; mid-range vs budget cost classes). They are not alternatives at one budget: pick Mercury 2 when you need intelligence above 8.5, and DeepSeek V3 (Dec '24) when 8.5 is enough.
| Metric | DeepSeek V3 (Dec '24) | Mercury 2 | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 8.5 | 10.8 | +27.1% Mercury 2 is 27.1% more intelligent |
| Total AA cost (lower is better) | $25.37 | $104.94 | +313.6% Mercury 2 is 4.1× more expensive |
| End-to-end time (lower is better) | 10.23s | 6.65s | −35.0% Mercury 2 is 35.0% faster |
| Price per 1M tokens (in / out) | $0.32 / $0.89 | $0.25 / $0.75 | |
| Input modalities | Text | Text |
Mercury 2 vs DeepSeek V3 (Dec '24): 27.1% more intelligence, 313.6% more cost to run the AA benchmark, and 35.0% faster end-to-end.
Today's Pareto frontier
Against the frontier
- DeepSeek V3 (Dec '24) is off the frontier: at the same or lower cost, GPT-6 Luna (Low) scores 21.50 for $10.6.
- Mercury 2 is off the frontier: at the same or lower cost, Claude Haiku 5.5 (High) scores 37.80 for $94.1.
DeepSeek V3 (Dec '24)Mercury 2Today'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-v3/mercury-2 · Markdown: /api/v1/compare/deepseek-v3/mercury-2?format=md · llms.txt · agent skill