Claude 3 Haiku vs Qwen3.8 27B Abliterated
Claude 3 Haiku is cheaper on output tokens, while Qwen3.8 27B Abliterated offers a larger context window. Choose Claude 3 Haiku or Qwen3.8 27B Abliterated based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Claude 3 Haiku | Qwen3.8 27B Abliterated |
|---|---|---|
| Provider | Anthropic | NEAR AI Cloud |
| Input / 1M tokens | $0.25 | $0.04 |
| Output / 1M tokens | $1.25 | $1.35 |
| Context window | 200K | 262K |
| Parameters | — | — |
| Open weights | — | No |
| Released | — | Oct 2026 |
| LMArena (Chatbot Arena) Elo score | 1195 | — |
| Epoch Capabilities Index score | 119 | — |
| SWE-bench Verified score | 40.6 | — |
| GPQA Diamond score | 36.3 | — |
| LMArena Vision Elo score | 950 | — |
FAQ
Is Claude 3 Haiku or Qwen3.8 27B Abliterated cheaper?
Claude 3 Haiku is cheaper on output tokens ($1.25 vs $1.35 per 1M).
Which has the larger context window, Claude 3 Haiku or Qwen3.8 27B Abliterated?
Qwen3.8 27B Abliterated has the larger context window (262K tokens).
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.