Qwen3-235B-A22B-Thinking-2507 is a high-performance, open-weight Mixture-of-Experts (MoE) language model optimized for complex reasoning tasks. It activates 22B of its 235B parameters per forward pass and natively supports up to 262,144...
Model details →Qwen: Qwen3 235B A22B Thinking 2507 vs Tencent: Hy3
Tencent: Hy3 wins on 2 of 6 axes — pricing and capability skew in its favour for most workloads.
Hy3 is a 295B-parameter Mixture-of-Experts model from Tencent (21B active, 192 experts with top-8 routing) built for reasoning, agentic workflows, and real-world production use. It supports a configurable reasoning effort:...
Model details →Side-by-side comparison
| Capability | Qwen: Qwen3 235B A22B Thinking 2507 | Tencent: Hy3 | Winner |
|---|---|---|---|
| Context window Maximum number of input tokens the model can attend to in a single request. | 262K | 262K | Tie |
| Input price (per 1M) Cost per million input tokens billed by the provider. | $0.23 | $0.13 | 🏆 Tencent: Hy3 |
| Output price (per 1M) Cost per million output tokens billed by the provider. | $2.30 | $0.53 | 🏆 Tencent: Hy3 |
| Tool / function calling First-class support for emitting structured tool calls. | Yes | Yes | Tie |
| Vision input Accepts image inputs alongside text. | — | — | Tie |
| Reasoning mode Internal chain-of-thought / extended-thinking support. | Yes | Yes | Tie |
Frequently asked questions
Is Qwen: Qwen3 235B A22B Thinking 2507 better than Tencent: Hy3?
Tencent: Hy3 wins on 2 of 6 axes — pricing and capability skew in its favour for most workloads.
What's the price difference between Qwen: Qwen3 235B A22B Thinking 2507 and Tencent: Hy3?
Input: $0.23 vs $0.13 per 1M tokens. Output: $2.30 vs $0.53 per 1M tokens.
What context windows do Qwen: Qwen3 235B A22B Thinking 2507 and Tencent: Hy3 support?
Qwen: Qwen3 235B A22B Thinking 2507 supports up to 262K tokens. Tencent: Hy3 supports up to 262K tokens.
Do both Qwen: Qwen3 235B A22B Thinking 2507 and Tencent: Hy3 support tool calling?
Qwen: Qwen3 235B A22B Thinking 2507: yes. Tencent: Hy3: yes.
Compare Qwen: Qwen3 235B A22B Thinking 2507 with other models
vs Anthropic Claude Haiku Latestvs Anthropic Claude Sonnet Latestvs Anthropic: Claude Fable Latestvs Anthropic: Claude Opus Latest