The AI arena is free today

Open Superagent

Codestral-22B vs Phi-3.5-mini-instruct

Codestral-22B leads the LLM Stats Score 0.1 to -3.8.

Mistral AI · Microsoft · Updated for 2026

Which is better?

Codestral-22B leads the overall LLM Stats Score 0.1 to -3.8, ranking #314 overall.

In the 2 individual benchmarks reported for both models, Codestral-22B wins 2; this is a narrower head-to-head signal than the composite indexes.

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

Choose Codestral-22B

  • overall performance matters — it scores 0.1 and ranks #314 on LLM Stats
  • your work emphasizes reasoning — it leads those capability indexes
  • you value its reported benchmark strengths — it wins 2 of 2 exact shared results

Choose Phi-3.5-mini-instruct

  • you want the most recent training data — it shipped Aug 2024

At a glance

The differences that matter most.

Codestral-22B
Mistral AI
Phi-3.5-mini-instruct
Microsoft
Core performance indexes
LLM Stats Score
0.1
#314
-3.8
#334
Reasoning
-0.0
#307
-4.7
#331
Coding
2.5
#218
-6.8
#253
Cost, coverage & limits
Benchmark wins
2 of 2
0 of 2
Input price
— / M
$0.10 / M
Output price
— / M
$0.10 / M
Context window
—
128,000

Individual benchmarks

7 reported for Codestral-22B · 31 for Phi-3.5-mini-instruct

2 shared

Codestral-22B outperforms in 2 benchmarks (HumanEval, MBPP), while Phi-3.5-mini-instruct is better at 0 benchmarks.

Codestral-22B significantly outperforms across most benchmarks.

Thu Sep 03 2026 • llm-stats.com

Human preference

Blind head-to-head votes and playground preference scores

Model Size

Parameter count comparison

18.4B diff

Codestral-22B has 18.4B more parameters than Phi-3.5-mini-instruct, making it 484.2% larger.

Mistral AI
Codestral-22B
22.2Bparameters
Microsoft
Phi-3.5-mini-instruct
3.8Bparameters
22.2B
Codestral-22B
3.8B
Phi-3.5-mini-instruct

Context Window

Maximum input and output token capacity

Only Phi-3.5-mini-instruct specifies input context (128,000 tokens). Only Phi-3.5-mini-instruct specifies output context (128,000 tokens).

Mistral AI
Codestral-22B
Input- tokens
Output- tokens
Microsoft
Phi-3.5-mini-instruct
Input128,000 tokens
Output128,000 tokens
Thu Sep 03 2026 • llm-stats.com

License

Usage and distribution terms

Codestral-22B is licensed under MNPL-0.1, while Phi-3.5-mini-instruct 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

Phi-3.5-mini-instruct

MIT

Open weights

Release Timeline

When each model was launched

Codestral-22B was released on 2024-05-29, while Phi-3.5-mini-instruct was released on 2024-08-23.

Phi-3.5-mini-instruct is 3 months newer than Codestral-22B.

Codestral-22B

May 29, 2024

2.3 years ago

Phi-3.5-mini-instruct

Aug 23, 2024

2.0 years ago

2mo 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 Phi-3.5-mini-instruct side-by-side, then vote on the output you prefer.

Codestral-22B
✓ Preferred
Phi-3.5-mini-instruct
Open in Playground

FAQ

Common questions about Codestral-22B vs Phi-3.5-mini-instruct.

Which is better, Codestral-22B or Phi-3.5-mini-instruct?

Codestral-22B leads the LLM Stats Score 0.1 to -3.8. Codestral-22B is made by Mistral AI and Phi-3.5-mini-instruct is made by Microsoft. The best choice depends on your use case — compare their capability indexes, individual benchmarks, pricing, and limits above.

How does Codestral-22B compare to Phi-3.5-mini-instruct in benchmarks?

Codestral-22B scores HumanEvalFIM-Average: 91.6%, HumanEval: 81.1%, MBPP: 78.2%, Spider: 63.5%, HumanEval-Average: 61.5%. Phi-3.5-mini-instruct scores GSM8k: 86.2%, ARC-C: 84.6%, RULER: 84.1%, PIQA: 81.0%, OpenBookQA: 79.2%.

What are the context window sizes for Codestral-22B and Phi-3.5-mini-instruct?

Codestral-22B supports an unknown number of tokens and Phi-3.5-mini-instruct supports 128K 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 Phi-3.5-mini-instruct?

Key differences include LLM Stats Score (0.1 vs -3.8), licensing (MNPL-0.1 vs MIT). See the full comparison above for benchmark-by-benchmark results.

Who makes Codestral-22B and Phi-3.5-mini-instruct?

Codestral-22B is developed by Mistral AI and Phi-3.5-mini-instruct is developed by Microsoft.

Related comparisons