46.7B
Parameters
MoE (12.9B active)
AIVory Smart Inference · Mistral
Open WeightMixtral 8x7B - the original open MoE.
The model that proved MoE works at scale for open-weight inference. 46.7B total, 12.9B active. Apache 2.0 license, 32K context, and still one of the cheapest models available.
At a glance
Mixtral 8x7B via Smart Inference.
Routed to the cheapest live provider on every request. Same model, lower bill.
32K
Context window
tokens per request
—
Cheapest input price
per 1M tokens via Smart Inference
—%
Cheaper
vs the priciest endpoint for the same model
Pricing
Mixtral 8x7B: price comparison.
Our cheapest routed price vs what you'd pay elsewhere.
Prices in USD per 1M input tokens · updated live from the router
Quick start
Use Mixtral 8x7B in three lines.
OpenAI-compatible. Swap your base_url, keep everything else.
from openai import OpenAI
client = OpenAI(
base_url="https://smart.aivory.net/v1",
api_key="sk-aivory-...",
)
resp = client.chat.completions.create(
model="mixtral-8x7b",
messages=[{"role": "user", "content": "Ship it."}],
)
print(resp.choices[0].message.content)import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://smart.aivory.net/v1",
apiKey: process.env.AIVORY_API_KEY,
});
const resp = await client.chat.completions.create({
model: "mixtral-8x7b",
messages: [{ role: "user", content: "Ship it." }],
});
console.log(resp.choices[0].message.content);curl https://smart.aivory.net/v1/chat/completions \
-H "Authorization: Bearer $AIVORY_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "mixtral-8x7b",
"messages": [{"role": "user", "content": "Ship it."}]
}'That's the whole migration. Get early access and try it →
Use cases
What Mixtral 8x7B is great at.
-
Budget-Conscious Production Workloads
Mixtral gives you 70B-class quality at 13B inference cost. For tasks like email classification, FAQ answering, and data normalization, it delivers good enough output at the lowest cost tier.
-
European Language Tasks
Mistral trained Mixtral with strong French, German, Spanish, and Italian coverage. If your product serves EU users and you need a cheap multilingual backbone, Mixtral is a solid default.
-
Legacy System Migration
Mixtral has been running in production since late 2023. If you're already using it elsewhere and want to cut costs, point your existing code at Smart Inference. Same model, cheaper rate, zero migration effort.
FAQ
Common questions.
How cheap is Mixtral 8x7B on AIVory?
Typically one of the lowest-cost models in the catalogue. The 12.9B active parameter count means providers can serve it on smaller GPUs, which drives the price down. Check the live table above.
Is Mixtral still competitive with newer models?
For simple tasks, yes. It's been outpaced by Llama 3.3 and Gemma 4 on reasoning benchmarks, but for classification, extraction, and short completions, it's hard to beat on price. Use newer models when quality matters more than cost.
What's the license?
Apache 2.0. The most permissive license of any model in the catalogue. No restrictions on commercial use, no MAU limits, no attribution requirements beyond the license itself.
Why is the context window only 32K?
Mixtral was released before the 128K context wave. 32K is enough for most single-document tasks but too short for multi-document analysis. If you need long context, look at Llama 3.3 70B or DeepSeek V3.2.
Start using Mixtral 8x7B. Today.
Swap your base_url, keep shipping.
The router is live. Credits from $10, no subscription, no commitment.