The AI arena is free today

Open Superagent

Codestral-22B vs DeepSeek-R1

Comparing Codestral-22B and DeepSeek-R1 across benchmarks, pricing, and capabilities.

Mistral AI · DeepSeek · Updated for 2026

Which is better?

Codestral-22B and DeepSeek-R1 trade strengths across price, capabilities, and technical limits. The better choice depends on the workload.

Based on current LLM Stats indexes, shared benchmarks, pricing, and model metadata for 2026.

Choose Codestral-22B

  • you are already invested in the Mistral AI ecosystem

Choose DeepSeek-R1

  • you want the most recent training data — it shipped Jan 2025

At a glance

The differences that matter most.

Benchmark wins
Input price
— / M
$0.55 / M
Output price
— / M
$2.19 / M
Context window
131,072

Individual benchmarks

7 reported for Codestral-22B · 0 for DeepSeek-R1

No common benchmarks found

Codestral-22B and DeepSeek-R1don't have any common benchmark datasets to compare. They may have been evaluated on different testing suites.

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

648.8B diff

DeepSeek-R1 has 648.8B more parameters than Codestral-22B, making it 2922.5% larger.

Mistral AI
Codestral-22B
22.2Bparameters
DeepSeek
DeepSeek-R1
671.0Bparameters
22.2B
Codestral-22B
671.0B
DeepSeek-R1

Context Window

Maximum input and output token capacity

Only DeepSeek-R1 specifies input context (131,072 tokens). Only DeepSeek-R1 specifies output context (131,072 tokens).

Mistral AI
Codestral-22B
Input- tokens
Output- tokens
DeepSeek
DeepSeek-R1
Input131,072 tokens
Output131,072 tokens
Mon Sep 07 2026 • llm-stats.com

License

Usage and distribution terms

Codestral-22B is licensed under MNPL-0.1, while DeepSeek-R1 uses MIT.

License differences may affect how you can use these models in commercial or open-source projects.

Codestral-22B

MNPL-0.1

Open weights

DeepSeek-R1

MIT

Open weights

Release Timeline

When each model was launched

Codestral-22B was released on 2024-05-29, while DeepSeek-R1 was released on 2025-01-20.

DeepSeek-R1 is 8 months newer than Codestral-22B.

Codestral-22B

May 29, 2024

2.3 years ago

DeepSeek-R1

Jan 20, 2025

1.6 years ago

7mo newer

Knowledge Cutoff

When training data ends

Neither model specifies a knowledge cutoff date.

Unable to compare the recency of their training data.

No cutoff dates available

Outputs Comparison

Notice missing or incorrect data?Start an Issue discussion

Judge for yourself.

Run your own prompts against Codestral-22B and DeepSeek-R1 side-by-side, then vote on the output you prefer.

Codestral-22B
✓ Preferred
DeepSeek-R1
Open in Playground

FAQ

Common questions about Codestral-22B vs DeepSeek-R1.

Which is better, Codestral-22B or DeepSeek-R1?

Codestral-22B (Mistral AI) and DeepSeek-R1 (DeepSeek) each have strengths in different areas. Compare their benchmark scores, pricing, context windows, and capabilities above to determine which fits your needs.

How does Codestral-22B compare to DeepSeek-R1 in benchmarks?

Codestral-22B scores HumanEvalFIM-Average: 91.6%, HumanEval: 81.1%, MBPP: 78.2%, Spider: 63.5%, HumanEval-Average: 61.5%.

What are the context window sizes for Codestral-22B and DeepSeek-R1?

Codestral-22B supports an unknown number of tokens and DeepSeek-R1 supports 131K tokens. A larger context window lets you process longer documents, conversations, or codebases in a single request.

What are the main differences between Codestral-22B and DeepSeek-R1?

Key differences include licensing (MNPL-0.1 vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Codestral-22B and DeepSeek-R1?

Codestral-22B is developed by Mistral AI and DeepSeek-R1 is developed by DeepSeek.