impulse

Qwen3 4B

Open weights
API id qwen3-4b-fp8
Input / 1M
$0.03
at Novita AI
Output / 1M
$0.03
at Novita AI
Context window
128K
tokens
Max output
8K
tokens per reply
Hosts
3
3 listings
Released
-
date not published

Pricing

List prices at Novita AI, in US dollars per million tokens.

TierInput / 1MOutput / 1M
Standard$0.03$0.03

What it costs a month

At Novita AI list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$11.25
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$15.48
  • Coding agent1,000 requests a day, 40K in / 2K out$37.80
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$79.20
Use your own numbers

Where to run it

3 hosts serve Qwen3 4B. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Novita AIInference
qwen3-4b-fp8
$0.03 / $0.03
Qwen3-4B
$0.08 / $0.24
+300% vs headline
Fireworks AIInference
qwen3-4b
$0.20 / $0.20
+567% vs headline

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-2.0
Runtimes: 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 Novita AI.

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

Model line

Newer and older versions, at each lab's own price.

  1. UndatedQwen3.5 4BLatest$0.04 / $0.07
  2. UndatedQwen3 4BThis page$0.03 / $0.03

Call it

Using Novita AI's API.

Python
from litellm import completion

response = completion(
    model="novita/qwen3-4b-fp8",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Qwen3 4B cost?

On Novita AI, Qwen3 4B costs $0.03 per million input tokens and $0.03 per million output tokens.

Which APIs offer Qwen3 4B?

Qwen3 4B runs on 3 APIs: Novita AI, Nebius AI Studio and Fireworks AI.

What is Qwen3 4B's context window?

128K tokens, with up to 8K tokens in one reply.

Is Qwen3 4B being retired?

No retirement date is scheduled for it on any API we track.

Is there a newer version of Qwen3 4B?

Yes. Qwen3.5 4B is the newest model in this line.

What are cheaper alternatives to Qwen3 4B?

Models that cost less and keep most of what it can do: gpt-oss-20b on Darkbloom (5% less). Savings are on a mix of 3 input tokens to 1 output token.

What is Qwen3 4B's API model id?

Use qwen3-4b-fp8 with Novita AI. Other hosts use their own ids: Qwen3-4B on Nebius AI Studio, qwen3-4b on Fireworks AI.

Sources

Raw catalog entry for qwen3-4b-fp8 on Novita AI
{
  "litellm_provider": "novita",
  "mode": "chat",
  "input_cost_per_token": 3e-8,
  "output_cost_per_token": 3e-8,
  "max_input_tokens": 128000,
  "max_output_tokens": 8192,
  "supports_system_messages": true,
  "supports_reasoning": true,
  "supports_function_calling": true,
  "supports_tool_choice": true
}