Phi 4 vs Mistral Large
Updated October 21, 2026. 34 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Neither model has a coding score yet, so the two cannot be ranked on coding.
Phi 4 has the edge where output tokens cost 42.9x less, and where what matters is LMArena Text: coding or LMArena Text: software and IT, where it leads.
- Shared results
- 34
- Phi 4 only
- 11
- Mistral Large only
- 10
- Like-for-like categories
- 6 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Phi 4
Phi 4 leads, 5.9 against 5.8.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only Mistral Large has a GPQA Diamond result, 47.7%.
Directional only
Math
AIME
No call
Only Mistral Large has a AIME result, 9.2%.
Directional only
Knowledge
MMLU Pro
No call
Only Mistral Large has a MMLU Pro result, 69.7%.
Directional only
Long documents
Listed context window
Mistral Large
Mistral Large takes 128K tokens against 16K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Phi 4
Phi 4 $0.0001, Mistral Large $0.0050 per turn.
List prices
At a glance
- Coding index
- Phi 4 not benchmarked, Mistral Large not benchmarked.
- Widest category
- Knowledge, 11.0 points apart on average over 6 shared benchmarks.
- Widest single gap
- GPQA Diamond (Artificial Analysis run): Phi 4 57.5% against Mistral Large 35.1%, 22.4 points apart on a 0 to 100 scale.
- Output price
- Phi 4 $0.14 against Mistral Large $6.00 per million output tokens, 42.9x apart.
- Context window
- Phi 4 takes 16K tokens, Mistral Large 128K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Phi 4 | Bars | Mistral Large | Ties |
|---|---|---|---|---|
| Coding | 1 | 2 | ||
| Intelligence | 2 | 0 | ||
| Reasoning | 2 | 3 | ||
| Math | 0 | 2 | ||
| Knowledge | 2 | 4 | ||
| General | 0 | 16 | ||
| All shared boards | 7 | 27 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index12
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| LiveCodeBench (Vals AI) | Not measured | 37.1% | Only Mistral Large measured |
| LiveCodeBench (Artificial Analysis run) | 23.1% | 17.8% | Phi 4 |
| Terminal-Bench Hard (Artificial Analysis run) | 3.8% | Not measured | Only Phi 4 measured |
| LMArena Text: coding | 1231 | 1275 | Mistral Large |
| LMArena Text: software and IT | 1228 | 1274 | Mistral Large |
| Neither model has a result on the other 33 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| AA Intelligence Index | 5.9 | 5.8 | Phi 4 |
| Epoch Capabilities Index | 130.5 | 122.1 | Phi 4 |
Agentic workDirectional onlyPublic boards, shown as published, kept out of the index2 boards, none shared
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 0.0% | Not measured | Only Phi 4 measured |
| BALROG | 11.6% | Not measured | Only Phi 4 measured |
| Neither model has a result on the other 37 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index23
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 47.7% | Only Mistral Large measured |
| GPQA Diamond (Artificial Analysis run) | 57.5% | 35.1% | Phi 4 |
| Humanity's Last Exam (Artificial Analysis run) | 3.8% | 3.5% | Phi 4 |
| GPQA Diamond (Epoch AI run) | 56.1% | Not measured | Only Phi 4 measured |
| LMArena Text: hard prompts | 1220 | 1257 | Mistral Large |
| LMArena Text: expert | 1203 | 1208 | Mistral Large |
| LMArena Text: hard prompts in English | 1230 | 1268 | Mistral Large |
| Neither model has a result on the other 14 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| AIME | Not measured | 9.2% | Only Mistral Large measured |
| MATH 500 (Vals AI) | Not measured | 74.4% | Only Mistral Large measured |
| MGSM (Vals AI) | Not measured | 87.2% | Only Mistral Large measured |
| AA Math Index | 18.0 | Not measured | Only Phi 4 measured |
| AIME 2025 (Artificial Analysis run) | 18.0% | Not measured | Only Phi 4 measured |
| OTIS Mock AIME | 13.8% | Not measured | Only Phi 4 measured |
| MATH level 5 | 64.9% | Not measured | Only Phi 4 measured |
| LMArena Text: math | 1246 | 1261 | Mistral Large |
| LMArena Text: mathematical work | 1254 | 1257 | Mistral Large |
| Neither model has a result on the other 4 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index24
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 69.7% | Only Mistral Large measured |
| TaxEval (Vals AI) | Not measured | 63.8% | Only Mistral Large measured |
| MedQA (Vals AI) | Not measured | 76.2% | Only Mistral Large measured |
| MMLU Pro (Artificial Analysis run) | 71.4% | 51.5% | Phi 4 |
| MMLU | 84.8% | 68.8% | Phi 4 |
| LMArena Text: science | 1206 | 1260 | Mistral Large |
| LMArena Text: medicine and healthcare | 1202 | 1255 | Mistral Large |
| LMArena Text: legal and government | 1241 | 1275 | Mistral Large |
| LMArena Text: business and finance | 1199 | 1238 | Mistral Large |
| Neither model has a result on the other 8 boards here | |||
Web designNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| Neither model has a result on 9 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index016
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| IFBench | 23.5% | Not measured | Only Phi 4 measured |
| Long context reasoning (AA-LCR) | 0.0% | Not measured | Only Phi 4 measured |
| DTBench | Not measured | 60.8% | Only Mistral Large measured |
| Creative writing (Lech Mazur) | 62.6% | Not measured | Only Phi 4 measured |
| LMArena Text | 1217 | 1265 | Mistral Large |
| LMArena Text: instruction following | 1201 | 1249 | Mistral Large |
| LMArena Text: creative writing | 1181 | 1242 | Mistral Large |
| LMArena Text: English | 1231 | 1284 | Mistral Large |
| LMArena Text: non-English | 1197 | 1235 | Mistral Large |
| LMArena Text: Chinese | 1212 | 1239 | Mistral Large |
| LMArena Text: German | 1222 | 1242 | Mistral Large |
| LMArena Text: Japanese | 1158 | 1170 | Mistral Large |
| LMArena Text: Korean | 1151 | 1202 | Mistral Large |
| LMArena Text: French | 1224 | 1325 | Mistral Large |
| LMArena Text: Spanish | 1234 | 1268 | Mistral Large |
| LMArena Text: Russian | 1209 | 1252 | Mistral Large |
| LMArena Text: Polish | Not measured | 1157 | Only Mistral Large measured |
| LMArena Text: multi-turn | 1206 | 1260 | Mistral Large |
| LMArena Text: longer queries | 1217 | 1261 | Mistral Large |
| LMArena Text: writing and language | 1198 | 1261 | Mistral Large |
| LMArena Text: entertainment and media | 1162 | 1230 | Mistral Large |
| Neither model has a result on the other 25 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Phi 4 | Mistral Large | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Phi 4 costs less
Repository review
50K input + 3K output tokens
Phi 4 costs less
Agent loop
200K input + 10K output tokens
Phi 4 costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Microsoft Phi 4 | Mistral Large |
|---|---|---|
| Maker | Microsoft | Mistral AI |
| Weights | Open weights | No public weights listed |
| Released | January 10, 2025 | February 26, 2024 |
| Context window | 16K | 128K |
| Input $/Mtok | $0.07 | $2.00 |
| Output $/Mtok | $0.14 | $6.00 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Phi 4 better than Mistral Large for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Phi 4 has the edge where output tokens cost 42.9x less, and where what matters is LMArena Text: coding or LMArena Text: software and IT, where it leads.
Which is cheaper, Phi 4 or Mistral Large?
Phi 4 costs $0.07 per million input tokens and $0.14 per million output tokens; Mistral Large costs $2.00 and $6.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Phi 4 or Mistral Large?
Mistral Large, with 128K tokens against 16K.
How many benchmarks is the Phi 4 vs Mistral Large comparison based on?
34 benchmarks have a result for both models as of October 21, 2026. Phi 4's coding index rests on 0 independent benchmarks and Mistral Large's on 0. Where only one model has a result, the page says so instead of naming a winner.