Qwen3 235B A22B Thinking 2507 (NovitaAI) vs Qwen3.8 27B (reka-ai)
Both have similar output pricing, while Qwen3.8 27B (reka-ai) offers a larger context window. Choose Qwen3 235B A22B Thinking 2507 (NovitaAI) or Qwen3.8 27B (reka-ai) based on the trade-off between cost, context, and the benchmarks that matter for your use case.
| Spec | Qwen3 235B A22B Thinking 2507 (NovitaAI) | Qwen3.8 27B (reka-ai) |
|---|---|---|
| Provider | LLM Gateway | LLM Gateway |
| Input / 1M tokens | $0.30 | $0.41 |
| Output / 1M tokens | $3.00 | $3.00 |
| Context window | 131K | 262K |
| Parameters | — | — |
| Open weights | No | Yes |
| Released | Jul 2025 | Aug 2026 |
FAQ
Pricing is indicative — confirm with the provider before production use. Highlighted values indicate the better figure for that row.