Qwen3 30B A3B Instruct 2507 is cheaper than Llama 3.1 405B at $0.2/M vs $2.7/M input tokens.
| Spec | Llama 3.1 405B | Qwen3 30B A3B Instruct 2507 |
|---|---|---|
| Input price | $2.7/M tokens | $0.2/M tokens |
| Output price | $3.1/M tokens | $0.8/M tokens |
| Context window | 128K tokens | 131K tokens |
| Released | Jul 2024 | Jul 2025 |
Not enough votes to call it. On the specs, Llama 3.1 405B has the edge: bigger model tier, major provider backing.
Qwen3 30B A3B Instruct 2507 costs 3.9x less per token.
Qwen3 30B A3B Instruct 2507 is cheaper on both — 14× input, 3.9× output
Qwen3 30B A3B Instruct 2507 uses 666.0x more bold
Qwen3 30B A3B Instruct 2507 is cheaper than Llama 3.1 405B at $0.2/M vs $2.7/M input tokens.
| Spec | Llama 3.1 405B | Qwen3 30B A3B Instruct 2507 |
|---|---|---|
| Input price | $2.7/M tokens | $0.2/M tokens |
| Output price | $3.1/M tokens | $0.8/M tokens |
| Context window | 128K tokens | 131K tokens |
| Released | Jul 2024 | Jul 2025 |
Not enough votes to call it. On the specs, Llama 3.1 405B has the edge: bigger model tier, major provider backing.
Qwen3 30B A3B Instruct 2507 costs 3.9x less per token.
Qwen3 30B A3B Instruct 2507 is cheaper on both — 14× input, 3.9× output
Qwen3 30B A3B Instruct 2507 uses 666.0x more bold