Qwen3 Max Thinking
Qwen3-Max-ThinkingPricing
List prices at DeepInfra, in US dollars per million tokens.
| Tier | Input / 1M | Output / 1M | |
|---|---|---|---|
| Standard | $1.20 | $6.00 | |
| Cached input 80% off input | $0.24 | 80% off input |
What it costs a month
At DeepInfra list prices, 30 days.
- Chat assistant5,000 requests a day, 2K in / 0.5K out, cached$724
- RAG over documents2,000 requests a day, 8K in / 0.6K out, cached$562
- Coding agent1,000 requests a day, 40K in / 2K out, cached$994
- Overnight batch20,000 requests a day, 4K in / 0.4K out$4,320
Where to run it
1 host serve Qwen3 Max Thinking. The comparison is on blended price, 3 parts input to 1 part output.
| Host | In / out per 1M | Model id | |
|---|---|---|---|
DeepInfraInference Qwen3-Max-Thinking | $1.20 / $6.00 | Qwen3-Max-Thinking |
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
- Features
- Prompt caching
- Open weights
- License: Apache-2.0Runtimes: 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 DeepInfra.
Smaller Alibaba Qwen models
No cheaper Alibaba Qwen model keeps its key capabilities.
Other frontier labs
- 17% less
- 38% less
- 48% less
- 92% less
Open-weight picks
- 6% less
- 28% less
- 64% less
- 67% less
Call it
Using DeepInfra's API.
from litellm import completion
response = completion(
model="deepinfra/Qwen3-Max-Thinking",
messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)Quick answers
How much does Qwen3 Max Thinking cost?
On DeepInfra, Qwen3 Max Thinking costs $1.20 per million input tokens and $6.00 per million output tokens. Cached input costs $0.24 per million tokens.
Which APIs offer Qwen3 Max Thinking?
Qwen3 Max Thinking runs on 1 API: DeepInfra.
What is Qwen3 Max Thinking's context window?
256K tokens, with up to 256K tokens in one reply.
Is Qwen3 Max Thinking being retired?
No retirement date is scheduled for it on any API we track.
What are cheaper alternatives to Qwen3 Max Thinking?
Models that cost less and keep most of what it can do: GPT-6.1 Terra on OpenAI (17% less), Gemini 3.8 Flash on Google AI Studio (38% less) and Qwen4 Max on Alibaba DashScope (6% less). Savings are on a mix of 3 input tokens to 1 output token.
What is Qwen3 Max Thinking's API model id?
Use Qwen3-Max-Thinking with DeepInfra.
Sources
Raw catalog entry for Qwen3-Max-Thinking on DeepInfra
{
"max_output_tokens": 256000,
"max_input_tokens": 256000,
"input_cost_per_token": 0.0000012,
"output_cost_per_token": 0.000006,
"cache_read_input_token_cost": 2.4e-7,
"supports_prompt_caching": true,
"litellm_provider": "deepinfra",
"mode": "chat",
"supports_function_calling": true,
"supports_tool_choice": true,
"supports_response_schema": true,
"supports_vision": false,
"source": "https://deepinfra.com/pricing"
}