impulse

Ernie 4.5 21B A3B Thinking

API id ernie-4.5-21B-a3b-thinking
Input / 1M
$0.07
at Novita AI
Output / 1M
$0.28
at Novita AI
Context window
131K
tokens
Max output
66K
tokens per reply
Hosts
1
1 listings
Released
-
date not published

Pricing

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

TierInput / 1MOutput / 1M
Standard$0.07$0.28

What it costs a month

At Novita AI list prices, 30 days.

  • Chat assistant5,000 requests a day, 2K in / 0.5K out$42.00
  • RAG over documents2,000 requests a day, 8K in / 0.6K out$43.68
  • Coding agent1,000 requests a day, 40K in / 2K out$101
  • Overnight batch20,000 requests a day, 4K in / 0.4K out$235
Use your own numbers

Where to run it

1 host serve Ernie 4.5 21B A3B Thinking. The comparison is on blended price, 3 parts input to 1 part output.

HostIn / out per 1M
Novita AIInference
ernie-4.5-21B-a3b-thinking
$0.07 / $0.28

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

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 Independent models

No cheaper Independent model keeps its key capabilities.

Other frontier labs

No cheaper match from another frontier lab.

Open-weight picks

  • Nemotron 3.5 Lightning
    NVIDIA · on Nebius AI Studio
    $0.06 / $0.24
    Has every key capability
    14% less
  • MiniMax M3.1 Flash
    MiniMax · on Refuseless
    $0.05 / $0.25
    Has every key capability
    18% less
  • Qwen3.5 9B
    Alibaba Qwen · on EmpirioLabs AI
    $0.09 / $0.13
    Has every key capability
    18% less
  • GLM 5.3 Flash
    Zhipu AI · on CLF AI Gateway
    $0.06 / $0.20
    Has every key capability
    22% less

Call it

Using Novita AI's API.

Python
from litellm import completion

response = completion(
    model="novita/ernie-4.5-21B-a3b-thinking",
    messages=[{"role": "user", "content": "Hello"}],
)
print(response.choices[0].message.content)

Quick answers

How much does Ernie 4.5 21B A3B Thinking cost?

On Novita AI, Ernie 4.5 21B A3B Thinking costs $0.07 per million input tokens and $0.28 per million output tokens.

Which APIs offer Ernie 4.5 21B A3B Thinking?

Ernie 4.5 21B A3B Thinking runs on 1 API: Novita AI.

What is Ernie 4.5 21B A3B Thinking's context window?

131K tokens, with up to 66K tokens in one reply.

Is Ernie 4.5 21B A3B Thinking being retired?

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

What are cheaper alternatives to Ernie 4.5 21B A3B Thinking?

Models that cost less and keep most of what it can do: Nemotron 3.5 Lightning on Nebius AI Studio (14% less). Savings are on a mix of 3 input tokens to 1 output token.

What is Ernie 4.5 21B A3B Thinking's API model id?

Use ernie-4.5-21B-a3b-thinking with Novita AI.

Sources

Raw catalog entry for ernie-4.5-21B-a3b-thinking on Novita AI
{
  "litellm_provider": "novita",
  "mode": "chat",
  "input_cost_per_token": 7e-8,
  "output_cost_per_token": 2.8e-7,
  "max_input_tokens": 131072,
  "max_output_tokens": 65536,
  "supports_system_messages": true,
  "supports_reasoning": true
}