qwen3-30b-a3b-instruct-2507
1383 ? Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference. It operates in non-thinking mode and is designed for high-quality instruction following, multilingual understanding, and...
LLM & Generative AI 1 model 1 listings snapshot Oct 14, 2026, 12:43 AM
Platforms
click a card to open the detail tableModel variants
W3r9 dedup — same model + identical tags merge; prices are listed per billing form| Model | Provider | Tags | Input /1M | Output /1M |
|---|---|---|---|---|
| qwen/qwen3-30b-a3b-instruct-2507 | OpenRouter | 30b256k context | $0.100 | $0.300 |
Capabilities
official design_arena & architecture metadataDesign Arena Elo official design_arena benchmark (OpenRouter)
No official Design Arena benchmark from this provider. Showing official architecture metadata only.
Input / Output modalities modality: text->text
Platform details
per provider| provider | model | Input /1M | Output /1M | Latency ? | Official $/1M | Availability | Data time |
|---|---|---|---|---|---|---|---|
| OpenRouter | qwen/qwen3-30b-a3b-instruct-2507 | $0.100 | $0.300 | 128.1 ms | — | 100% available | 2026-10-14 00:43:21 |
| Together | Qwen/Qwen3-30B-A3B-Instruct-2507-Lora | $0 | $0 | — | — | — | 2026-10-14 00:53:13 |
2 rows · snapshot 2026-10-14 00:43:21 · from data/public/prices.json + measurements.json