Inkling Small vs Qwen3.8 27B (Xhigh)
Which is better for the money: Inkling Small by Thinking Machines or Qwen3.8 27B (Xhigh) 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 27B (Xhigh) scores 8.0 points higher (33.7 vs 25.7) but costs 5.5× as much to run the benchmark ($1336 vs $241; premium vs mid-range cost classes). They are not alternatives at one budget: pick Qwen3.8 27B (Xhigh) when you need intelligence above 25.7, and Inkling Small when 25.7 is enough.
| Metric | Inkling Small | Qwen3.8 27B (Xhigh) | Difference |
|---|---|---|---|
| Intelligence (higher is better) | 25.7 | 33.7 | +31.1% Qwen3.8 27B (Xhigh) is 31.1% more intelligent |
| Total AA cost (lower is better) | $240.85 | $1,336.09 | +454.7% Qwen3.8 27B (Xhigh) is 5.5× more expensive |
| End-to-end time (lower is better) | 12.80s | 60.11s | +369.6% Qwen3.8 27B (Xhigh) is 369.6% slower |
| Price per 1M tokens (in / out) | $0.30 / $1.20 | $0.50 / $3.00 | |
| Input modalities | Text, Image, Speech | Text, Image, Video |
Qwen3.8 27B (Xhigh) vs Inkling Small: 31.1% more intelligence, 454.7% more cost to run the AA benchmark, and 369.6% slower end-to-end.
Today's Pareto frontier
Against the frontier
- Inkling Small is off the frontier: at the same or lower cost, MiMo-V2.6-Pro scores 46.30 for $207.
- Qwen3.8 27B (Xhigh) is off the frontier: at the same or lower cost, GPT-6.1 Sol (Max) scores 51.80 for $1082.
Inkling SmallQwen3.8 27B (Xhigh)Today'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/inkling-small/qwen3-8-27b · Markdown: /api/v1/compare/inkling-small/qwen3-8-27b?format=md · llms.txt · agent skill