Qwen3 30B A3B Instruct 2507
Qwen3-30B-A3B-Instruct-2507 is a 30.5B-parameter mixture-of-experts language model from Qwen, with 3.3B active parameters per inference.
Qwen3-30B-A3B-Instruct-2507Released 29 Jul 2025 (Jul 2025)Pricing
List prices at Nebius AI Studio, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $0.10 | $0.30 | |
| Cached input | $0.10 |
What it costs a month
At Nebius AI Studio list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$52.50
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$58.80
- Coding agent1,000 requests a day, 40K in / 2K out, cached$138
- Overnight batch20,000 requests a day, 4K in / 0.4K out$312
Where to run it
3 hosts serve Qwen3 30B A3B Instruct 2507. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
Nebius AI StudioInference Qwen3-30B-A3B-Instruct-2507 | $0.10 / $0.30 | Qwen3-30B-A3B-Instruct-2507 | |
W&B InferenceInference Qwen3-30B-A3B-Instruct-2507 | $0.10 / $0.30 | Qwen3-30B-A3B-Instruct-2507 | |
Fireworks AIInference qwen3-30b-a3b-instruct-2507 | $0.50 / $0.50 +233% vs headline | qwen3-30b-a3b-instruct-2507 |
Capabilities
Core
- Streaming
- Function calling
- JSON mode
- Structured outputs
Multimodal
- Vision
- Image input
- Audio
- Image generation
- Video
Advanced
- Reasoning
- Tool calling
- MCP
- Prompt caching
- Batch API
- Embeddings
- Takes in
- Text
- Produces
- Text
- Open weights
- License: Apache-2.0Runs in about 73.3 GB of GPU memoryRuntimes: Ollama, SGLang, llama.cpp, vLLM
Cheaper alternatives
Current models that cost less and keep most of what this one can do: tools, vision, reasoning, structured output, caching and enough context.
Same model, cheaper host
No host sells it for less than Nebius AI Studio.
Smaller Alibaba Qwen models
No cheaper Alibaba Qwen model keeps its key capabilities.
Other frontier labs
No cheaper match from another frontier lab.
Open-weight picks
- 0% less
- 5% less
- 7% less
- 8% less
Call it
Using Nebius AI Studio's API.
from litellm import completion
response = completion(
model="nebius/Qwen3-30B-A3B-Instruct-2507",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Qwen3 30B A3B Instruct 2507 cost?
On Nebius AI Studio, Qwen3 30B A3B Instruct 2507 costs $0.10 per million input tokens and $0.30 per million output tokens. Cached input costs $0.10 per million tokens.
Which APIs offer Qwen3 30B A3B Instruct 2507?
Qwen3 30B A3B Instruct 2507 runs on 3 APIs: Nebius AI Studio, W&B Inference and Fireworks AI.
What is Qwen3 30B A3B Instruct 2507's context window?
262K tokens, with up to 262K tokens in one reply.
Is Qwen3 30B A3B Instruct 2507 being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to Qwen3 30B A3B Instruct 2507?
Models that cost less and keep most of what it can do: Ministral 8B 2512 on Mistral AI (0% less). Savings are on a mix of 3 input tokens to 1 output token.
When was Qwen3 30B A3B Instruct 2507 released?
Alibaba Qwen released it on 29 Jul 2025.
What is Qwen3 30B A3B Instruct 2507's API model id?
Use Qwen3-30B-A3B-Instruct-2507 with Nebius AI Studio. Other hosts use their own ids: qwen3-30b-a3b-instruct-2507 on Fireworks AI.
Sources
Raw catalog entry for Qwen3-30B-A3B-Instruct-2507 on Nebius AI Studio
{
"max_input_tokens": 262144,
"max_output_tokens": 262144,
"input_cost_per_token": 1e-7,
"output_cost_per_token": 3e-7,
"litellm_provider": "nebius",
"mode": "chat",
"supports_function_calling": true,
"source": "https://tokenfactory.nebius.com/api/public/models_info",
"cache_read_input_token_cost": 1e-7
}