Qwen3.5 9B (Reasoning) vs Qwen3.8-Flash-Next
Which is better for the money: Qwen3.5 9B (Reasoning) by Alibaba or Qwen3.8-Flash-Next by Alibaba? Intelligence, total benchmark cost, and end-to-end time from Artificial Analysis & other sources, snapshot 2026-10-26.
Trade-off
Verdict
Depends on your budget
Qwen3.8-Flash-Next scores 28.6 points higher (39.8 vs 11.2) but costs 1.1× as much to run the benchmark ($363 vs $329; both mid-range class). Pick Qwen3.8-Flash-Next when the extra intelligence is worth the premium, otherwise Qwen3.5 9B (Reasoning).
| Metric | Qwen3.5 9B (Reasoning) | Qwen3.8-Flash-Next | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 11.2 | 39.8 | +255.4% Qwen3.8-Flash-Next is 3.6× more intelligent |
| Total AA cost (lower is better) | $328.86 | $362.72 | +10.3% Qwen3.8-Flash-Next is 10.3% more expensive |
| End-to-end time (lower is better) | 43.82s | 50.47s | +15.2% Qwen3.8-Flash-Next is 15.2% slower |
| Price per 1M tokens (in / out) | $0.14 / $0.20 | $0.15 / $0.47 | |
| Input modalities | Text, Image, Video | Text, Image, Video |
Qwen3.8-Flash-Next vs Qwen3.5 9B (Reasoning): 255.4% more intelligence, 10.3% more cost to run the AA benchmark, and 15.2% slower end-to-end.
Today's Pareto frontier
Against the frontier
- Qwen3.5 9B (Reasoning) is off the frontier: at the same or lower cost, MiMo-V2.6-Pro scores 46.30 for $207.
- Qwen3.8-Flash-Next is off the frontier: at the same or lower cost, GPT-6.1 Sol (Medium) scores 47.80 for $361.
Qwen3.5 9B (Reasoning)Qwen3.8-Flash-NextToday'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/qwen3-5-9b/qwen3-8-flash-next · Markdown: /api/v1/compare/qwen3-5-9b/qwen3-8-flash-next?format=md · llms.txt · agent skill