Live survey286 models · 13 providers · 17,592 auditsAudited Using Git Worktrees · SAFE →What changedHow we verifyllms.txt

Model comparison · Updated 17 Aug 2026

Phi 4 vs Mistral Small 3

Mistral Small 3 is about 1.4× cheaper on input ($0.05 against $0.07 per 1M tokens) and on output. Mistral Small 3 has the larger context window (33K against 16K tokens). For a support chatbot at 10,000 chats a day, the standard-tier bill is about $44.10 a month for Phi 4 and $29.70 for Mistral Small 3.

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
CoCopilotAIGoogle AI Mode
Cite

Cite this page

ModelsAtlas. “Phi 4 vs Mistral Small 3”. Retrieved 17 Aug 2026. https://modelsatlas.com/compare/microsoft/phi-4/vs/mistralai/mistral-small-24b-instruct

Model A

Phi 4

by Microsoft · weights not marked open · GRIDB1

Input / 1M
$0.07
observed We measured
Source
Our catalogue
Retrieved
17 Aug 2026
Output / 1M$0.14
Context16K
vs

Model B

Mistral Small 3

by Mistral AI · weights not marked open · GRIDB2

Input / 1M
$0.05
observed We measured
Source
Our catalogue
Retrieved
17 Aug 2026
Output / 1M$0.08
Context33K

Key numbers

Standard tier · highlighted = better

Phi 4
Mistral Small 3
Input / 1M tokens
$0.07
$0.05
Output / 1M tokens
$0.14
$0.08
Batch input / output
Not listed
Not listed
Context window
16,384
32,768
Input types
Not disclosed
Not disclosed
Weights
Not marked open
Not marked open
Tiers listed
standard
standard
Prices checked
2026-08-17
2026-08-17

What it costs for real workloads

Monthly, standard tier, 30 days

Support chatbot10k chats/day · 1.5k in, 300 out
$44.10/mo
$29.70/mo
RAG over documents2k queries/day · 12k in, 500 out
$54.60/mo
$38.40/mo
Coding agent300 tasks/day · 60k in, 8k out
$47.88/mo
$32.76/mo

Phi 4   Mistral Small 3 · Derived from the standard-tier prices above, input and output tokens per request as stated, 30 days. Excludes caching discounts and batch. Try your own numbers.

On the atlas

Price band (A–H) × context band (1–6)

1 square apart. Mistral Small 3 (B2) sits 1 context band larger than Phi 4 (B1).

Same square as Mistral Small 3: Llama 3.2 1B Instruct ($0.03), Command R7B (12-2024) ($0.04). Open B1 · Open B2

AFreeB<$0.10C$0.10–0.30D$0.30–1E$1–3F$3–10G$10–30H$30+1<32K232–128K3128–256K4256K–1M51M6>1MINPUT PRICE PER 1M TOKENS →CONTEXT WINDOW ↑ · position within a square follows the real valueCommand R7B (12-2024) · B2Llama 3.2 1B Instruct · B2gpt-oss-safeguard-20b · B3Gemma 3 27B · B3Qwen3 32B · B3Gemma 3 12B · B3Gemma 3 4B · B3Llama 3.1 8B Instruct · B3Llama 3.2 3B Instruct · B3Phi 4Phi 4Mistral Small 3Mistral Small 3
Phi 4Mistral Small 3Nearest neighboursOpen weights
View the map as a table (3 squares)
SquarePrice bandContext bandModels
GRIDB1<$0.10<32KPhi 4
GRIDB2<$0.1032–128KMistral Small 3, Command R7B (12-2024), Llama 3.2 1B Instruct
GRIDB3<$0.10128–256Kgpt-oss-safeguard-20b, Gemma 3 27B, Qwen3 32B, Gemma 3 12B, Gemma 3 4B, Llama 3.1 8B Instruct, Llama 3.2 3B Instruct

Capabilities side by side

Only what our records list

Phi 4
Mistral Small 3
Vision input
Not disclosed
Not disclosed
Audio input
Not disclosed
Not disclosed
Open weights
Not marked open
Not marked open
Batch tier
Not listed
Not listed
Free tier
Not listed
Not listed
Thinking tier
Not listed
Not listed

Neither record lists input types, so we do not claim any.

Which should you choose?

Editorial

Choose Phi 4 if…

Nothing in the recorded numbers favours Phi 4 over Mistral Small 3. Pick it for what our records do not cover — your own tests, or an account you already have with Microsoft.

Choose Mistral Small 3 if…

  • you pay by the token: it costs $0.05 per 1M input against $0.07
  • output dominates your bill: $0.08 per 1M output against $0.14
  • your prompts are long: it takes 32,768 tokens against 16,384

Each reason is a fact from the tables above. Picking between them is our judgement, and we do not benchmark quality: price and context are what we can measure, so test both on your own prompts before you commit.

Questions

Is Phi 4 better than Mistral Small 3?
We do not benchmark quality, so we cannot say. On the recorded numbers: Mistral Small 3 is about 1.4× cheaper on input ($0.05 against $0.07 per 1M tokens) and on output. Mistral Small 3 has the larger context window (33K against 16K tokens). For a support chatbot at 10,000 chats a day, the standard-tier bill is about $44.10 a month for Phi 4 and $29.70 for Mistral Small 3. Which is better for you depends on your task; test both on your own prompts.
Which is cheaper, Phi 4 or Mistral Small 3?
Mistral Small 3: $0.05 against $0.07 per 1M input tokens, and $0.08 against $0.14 per 1M output, on standard pricing.
How much does each cost per 1M tokens?
Phi 4: $0.07 input and $0.14 output. Mistral Small 3: $0.05 input and $0.08 output. Standard tier, USD.
Which has the larger context window?
Mistral Small 3, with 32,768 tokens. Phi 4 has 16,384 tokens.
Do both offer batch pricing?
Neither lists a batch tier in our catalogue.
Can both read images or audio?
Input types recorded for Phi 4: none recorded. For Mistral Small 3: none recorded. Where a record lists none we say so rather than guess; check each vendor's documentation before relying on a capability.
Is either model open-weight?
Phi 4: not marked open-weight in our records. Mistral Small 3: not marked open-weight in our records.
What would each cost for a support chatbot?
At 10,000 chats a day, 1,500 input and 300 output tokens each, 30 days, standard tier: about $44.10 a month for Phi 4 and $29.70 for Mistral Small 3. This is our calculation from the listed prices, not a quote.

About the two models

Phi 4

By Microsoft. $0.07 input and $0.14 output per 1M tokens, 16K context, atlas square B1. Tiers listed: standard.

Full specs & pricing →

Mistral Small 3

By Mistral AI. $0.05 input and $0.08 output per 1M tokens, 33K context, atlas square B2. Tiers listed: standard.

Full specs & pricing →

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.