Mixtral 8x22B Instruct

Provider: Mistral AI EU Official API price

Input price
$2.00
/ 1M tokens
Output price
$6.00
/ 1M tokens
Cache price
$0.20
/ 1M tokens (read)
Context
66K tokens
Max output 52K tokens

Overview

Mixtral 8x22B Instruct is Mistral AI's flagship offering, designed for advanced language tasks requiring high performance. It provides robust capabilities for developers needing structured outputs and substantial context processing, positioning it as a reasoning specialist model.

Strengths

  • Supports reliable function-calling for integrating with external tools and APIs.
  • Offers JSON-mode for highly structured and easily parsable outputs.
  • Provides a competitive input token price of $2/1M compared to peers.
  • Features a significantly lower output token price of $6/1M than similar models.
  • Manages substantial conversational and document context with 65536 tokens.

Best use cases

  • Developing AI agents and assistants with precise tool-use capabilities.
  • Building applications requiring reliable structured data extraction into JSON format.
  • Powering advanced conversational interfaces that demand high quality and moderate context.
  • Integrating with systems where the cost of generating long outputs is a key consideration.
  • Automating complex workflows that rely on specific function execution.

Pricing position

Mixtral 8x22B Instruct is positioned as Mistral AI's most premium model, with input at $2/1M and output at $6/1M, making it significantly more expensive than siblings like Mistral Nemo. Against similarly priced peers like GPT-4.1 or Gemini 3.1 Pro Preview, its $2/1M input cost is competitive, while its $6/1M output cost is notably lower, as peers range from $8/1M to $12/1M. The cache read price is also very low at $0.2/1M tokens.

Caveats

The maximum output token limit for Mixtral 8x22B Instruct is currently unknown, which could impact use cases requiring very long generations. While its 65536-token context window is substantial, some Mistral AI sibling models, such as Mistral Nemo or Mistral Small 3.2 24B, offer larger contexts up to 131072 tokens for handling extremely long inputs. For tasks not requiring its advanced capabilities or highest performance, more cost-optimized Mistral models are available.

Notes drafted by an LLM from the figures on this page (2026-05-21) and rewritten when the price changes.

Key facts

Price basis
Official API price
Max output
52K tokens
Knowledge cutoff
2024-01-31
Listed on OpenRouter
2024-04-17
Tracked since
2026-05-08
Last updated
2026-11-10

Features

Function callingJSON output

Monthly cost estimate

Scale Input (M) Output (M) Monthly
0.1M (hobby) 0.1 0.05 $0.50
1M (small app) 1 0.5 $5.00
10M (mid scale) 10 5 $50.00
100M (large) 100 50 $500.00

※ Cache and batch discounts not factored in. Detailed calculator

Official resources

Other models from Mistral AI

Similar-priced models

Source: OpenRouter API (mistralai/mixtral-8x22b-instruct)