Mistral: Pixtral 12B

Mistral vision-language model for image understanding and multimodal chat

Model lab: Mistral
Model family: pixtral

Model resources

Hugging Face

Modalitiestext, image → text

Context window128,000

Maximum output128,000

Input · Context ≤ 128000 tokens · Maximum output 4000 tokensUSD 0.12 / 1M tokens

Output · Context ≤ 128000 tokens · Maximum output 4000 tokensUSD 0.12 / 1M tokens

Try now

Quickstart

Use your Moduck API Key, set the API base URL to https://api.moduck.ai/v1, and set the model parameter to mistral/pixtral-12b.

import os
import requests

base_url = "https://api.moduck.ai/v1"
api_key = os.environ["MODUCK_API_KEY"]
model_id = "mistral/pixtral-12b"

response = requests.post(
    f"{base_url}/chat/completions",
    headers={
        "Authorization": f"Bearer {api_key}",
        "Content-Type": "application/json",
    },
    json={
        "model": model_id,
        "messages": [{"role": "user", "content": "Hello!"}],
        "stream": False,
    },
    timeout=60,
)
response.raise_for_status()
print(response.json())

Use the familiar OpenAI-style chat request format to make model integration easier.

Pixtral 12B API Pricing

Type	Price information
Input · Context ≤ 128000 tokens · Maximum output 4000 tokens	USD 0.12 / 1M tokens
Output · Context ≤ 128000 tokens · Maximum output 4000 tokens	USD 0.12 / 1M tokens
Cached input · Context ≤ 128000 tokens · Maximum output 4000 tokens	USD 0.12 / 1M tokens
Cache write · Context ≤ 128000 tokens · Maximum output 4000 tokens	USD 0.12 / 1M tokens

Performance

Performance trends

Throughput87.8 tok/sCurrent

Time to first token1.4 sCurrent

End-to-end latency7.0 sCurrent

Throughput

Current87.8 tok/s

Window average89.7 tok/s

Time to first token

Current1.4 s

Window average1.4 s

End-to-end latency

Current7.0 s

Window average7.5 s

Tail latency (P99)

Current14.0 s

Window average16.8 s

Prompt cache hit rate

Current31.4%

Window average22.5%

Error rate

Current0.0%

Window average0.1%

Availability

Uptime100%At least one provider responded

Availability100%Requests actually served

Error rate0%Share of failed requests

Served requests, hour by hour

24 hours agoNow

Pixtral 12B Compare models

Model	Input price	Output price	Context window	Capabilities
Pixtral 12BThis page	USD 0.12 / 1M tokens · Context ≤ 128000 tokens · Maximum output 4000 tokens	USD 0.12 / 1M tokens · Context ≤ 128000 tokens · Maximum output 4000 tokens	128,000	Text output · Tool calling
Pixtral Large (25.02)	USD 1.5944 / 1M tokens	USD 4.7824 / 1M tokens	128,000	Text output · Tool calling
Pixtral Large (latest)	USD 1.6 / 1M tokens	USD 4.8 / 1M tokens	128,000	Text output · Tool calling
Mistral Large 4	USD 0.544 / 1M tokens	USD 1.672 / 1M tokens	1,048,576	Text output · Reasoning · Tool calling
Mistral Medium (latest)	USD 0.32 / 1M tokens	USD 1.6 / 1M tokens	262,144	Text output · Reasoning · Tool calling

Model specifications

Input: text, image
Output: text
Context window: 128,000
Maximum output: 128,000
Release date: 2024-09-01
Attachments: Yes
Reasoning: No
Tool calling: Yes
Temperature setting: Yes
Open weights: Yes

Frequently asked questions

What is Pixtral 12B for?

Mistral vision-language model for image understanding and multimodal chat

What inputs and outputs does Pixtral 12B support?

Recorded inputs: text, images; outputs: text.

How is Pixtral 12B priced?

The pricing table shows the current recorded Moduck price. For example: Input · Context ≤ 128000 tokens · Maximum output 4000 tokens USD 0.12 / 1M tokens; Output · Context ≤ 128000 tokens · Maximum output 4000 tokens USD 0.12 / 1M tokens. See the pricing table for all units, variants, and conditions.

What specifications are recorded for Pixtral 12B?

Current model records list Context window 128,000 tokens; Maximum output 128,000 tokens.

What model parameter does the Quickstart use?

The example uses mistral/pixtral-12b. This value comes from the page identifier and is not confirmed as a callable Moduck public ID.

Can I call this model through the Moduck API now?

This page has no verified Moduck API call for this model. A Quickstart example or reserved ID does not establish access; check the models available to your account.