Grok 4.3 (Non-reasoning) vs Qwen3.8 Max
Which is better for the money: Grok 4.3 (Non-reasoning) by SpaceXAI or Qwen3.8 Max by Alibaba? 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
Qwen3.8 Max scores 26.2 points higher (40.2 vs 14.0) but costs 13× as much to run the benchmark ($3172 vs $250; flagship vs mid-range cost classes). They are not alternatives at one budget: pick Qwen3.8 Max when you need intelligence above 14.0, and Grok 4.3 (Non-reasoning) when 14.0 is enough.
| Metric | Grok 4.3 (Non-reasoning) | Qwen3.8 Max | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 14.0 | 40.2 | +187.1% Qwen3.8 Max is 2.9× more intelligent |
| Total AA cost (lower is better) | $250.37 | $3,172.37 | +1167.1% Qwen3.8 Max is 13× more expensive |
| End-to-end time (lower is better) | 5.89s | 62.87s | +967.4% Qwen3.8 Max is 967.4% slower |
| Price per 1M tokens (in / out) | $1.25 / $2.50 | $2.00 / $6.00 | |
| Input modalities | Text, Image | Text, Image, Video |
Qwen3.8 Max vs Grok 4.3 (Non-reasoning): 187.1% more intelligence, 1167.1% more cost to run the AA benchmark, and 967.4% slower end-to-end.
Today's Pareto frontier
Against the frontier
- Grok 4.3 (Non-reasoning) is off the frontier: at the same or lower cost, MiMo-V2.6-Pro scores 46.30 for $207.
- Qwen3.8 Max is off the frontier: at the same or lower cost, Kimi K3.1 Pro scores 55.10 for $3117.
Grok 4.3 (Non-reasoning)Qwen3.8 MaxToday's Pareto frontierEvery other model
Matchup
Compare something else
Agents:
Accept: text/markdown or ?format=md returns this page as Markdown. JSON: /api/v1/compare/grok-4-3-non-reasoning/qwen3-8-max-0803 · Markdown: /api/v1/compare/grok-4-3-non-reasoning/qwen3-8-max-0803?format=md · llms.txt · agent skill