Qwen3.8 27B vs Grok 4.20 (Reasoning)
Qwen3.8 27B is cheaper on output tokens. Choose Qwen3.8 27B or Grok 4.20 (Reasoning) based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Qwen3.8 27B | Grok 4.20 (Reasoning) |
|---|---|---|
| Provider | Qwen | xAI |
| Input / 1M tokens | $0.04 | $1.25 |
| Output / 1M tokens | $1.25 | $2.50 |
| Context window | 1M | 1M |
| Parameters | — | — |
| Open weights | — | No |
| Released | — | Mar 2026 |
| LMArena (Chatbot Arena) Elo score | 1440 | — |
| LMArena WebDev Elo score | 1593 | — |
| Epoch Capabilities Index score | 152 | — |
| LMArena Vision Elo score | 1269 | — |
| LiveBench score | — | 69.0 |
FAQ
Is Qwen3.8 27B or Grok 4.20 (Reasoning) cheaper?
Qwen3.8 27B is cheaper on output tokens ($1.25 vs $2.50 per 1M).
Which has the larger context window, Qwen3.8 27B or Grok 4.20 (Reasoning)?
Both models offer a similar context window.
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.