Gemini 4 Carbon vs Mistral Large 4
Gemini 4 Carbon leads Mistral Large 4 on coding, 86.6 to 60.9 on the Vibe Coding Index, and Mistral Large 4 costs 4.8x less per output token ($2.10 vs $10.00 per million). Updated November 14, 2026; 60 benchmarks have a result for both.
Gemini 4 Carbon or Mistral Large 4: which to pick
Pick Gemini 4 Carbon for coding: it scores 86.6 on the Vibe Coding Index against 60.9 for Mistral Large 4, though on 4 benchmarks against 5.
Pick Mistral Large 4 for output tokens 4.8x cheaper ($2.10 vs $10.00 per million).
- Shared results
- 60
- Gemini 4 Carbon only
- 13
- Mistral Large 4 only
- 30
- Like-for-like categories
- 8 of 9
Only benchmarks that tested both models count toward a pick; a result only one model has is shown but not compared.
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
Gemini 4 Carbon
Gemini 4 Carbon scores 86.6 against 60.9.
4 and 5 benchmarks behind the scores
Intelligence
AA Intelligence Index
Gemini 4 Carbon
Gemini 4 Carbon leads, 60.0 against 26.1.
Same board, both measured
Web design
Design Arena: websites
Gemini 4 Carbon
Gemini 4 Carbon leads, 1319 against 1258.
Same board, both measured
General
Vals Index
Gemini 4 Carbon
Gemini 4 Carbon leads, 70.5% against 56.5%.
Same board, both measured
Long documents
Listed context window
No call
Both take 1.05M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Mistral Large 4
Gemini 4 Carbon $0.0070, Mistral Large 4 $0.0018 per turn.
List prices
At a glance
- Coding index
- Gemini 4 Carbon 86.6 from 4 benchmarks, Mistral Large 4 60.9 from 5 benchmarks.
- Widest category
- Math, 36.2 points apart on average over 5 shared benchmarks.
- Widest single gap
- FrontierMath, tier 4: Gemini 4 Carbon 95.1% against Mistral Large 4 26.8%, 68.3 points apart on a 0 to 100 scale.
- Output price
- Gemini 4 Carbon $10.00 against Mistral Large 4 $2.10 per million output tokens, 4.8x apart.
- Context window
- Both take 1.05M tokens.
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 | Gemini 4 Carbon | Bars | Mistral Large 4 | Ties |
|---|---|---|---|---|
| Coding | 11 | 0 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 6 | 0 | 1 tie | |
| Reasoning | 9 | 0 | ||
| Math | 5 | 0 | ||
| Knowledge | 6 | 0 | ||
| Web design | 6 | 0 | ||
| General | 14 | 0 | ||
| All shared boards | 59 | 0 | 1 tie |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index110
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| LiveBench coding | 91.0% | 79.4% | Gemini 4 Carbon |
| LMArena WebDev | 1768 | 1564 | Gemini 4 Carbon |
| Terminal-Bench 4 (Vals AI) | 63.1% | 27.8% | Gemini 4 Carbon |
| Vibe Code Bench (Vals AI) | 89.1% | 81.6% | Gemini 4 Carbon |
| Code migration (Vals AI) | Not measured | 30.6% | Only Mistral Large 4 measured |
| IOI (Vals AI) | Not measured | 45.3% | Only Mistral Large 4 measured |
| SWE-bench Multimodal (Vals AI) | Not measured | 33.1% | Only Mistral Large 4 measured |
| Vibe Code Bench 1-100 (Vals AI) | Not measured | 14.3% | Only Mistral Large 4 measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 65.2% | 24.7% | Gemini 4 Carbon |
| FrontierCode 2.0 | 40.6% | 15.0% | Gemini 4 Carbon |
| FrontierCode | Not measured | 39.4% | Only Mistral Large 4 measured |
| FrontierSWE | 61.2% | Not measured | Only Gemini 4 Carbon measured |
| MirrorCode | Not measured | 40.0% | Only Mistral Large 4 measured |
| ALE-Bench | 2374.7 | 1213.2 | Gemini 4 Carbon |
| LMArena Text: coding | 1564 | 1457 | Gemini 4 Carbon |
| LMArena Text: software and IT | 1552 | 1458 | Gemini 4 Carbon |
| LMArena WebDev: React | 1792 | 1570 | Gemini 4 Carbon |
| LMArena WebDev: HTML | 1770 | 1549 | Gemini 4 Carbon |
| GameBench 2 | Not measured | 57.6 | Only Mistral Large 4 measured |
| Terminal-Bench 4.0 | Not measured | 27.9% | Only Mistral Large 4 measured |
| CursorBench 4.0 | 60.4% | Not measured | Only Gemini 4 Carbon measured |
| Neither model has a result on the other 31 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| AA Intelligence Index | 60.0 | 26.1 | Gemini 4 Carbon |
| Epoch Capabilities Index | 169.8 | 154.6 | Gemini 4 Carbon |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| Design Arena: agents building mobile apps | 1284 | Not measured | Only Gemini 4 Carbon measured |
| Design Arena: spreadsheets | 1224 | Not measured | Only Gemini 4 Carbon measured |
| LiveBench research | 92.5% | 80.3% | Gemini 4 Carbon |
| LiveBench browser | 80.0% | 17.4% | Gemini 4 Carbon |
| LiveBench long agentic | 72.7% | 24.7% | Gemini 4 Carbon |
| Finance Agent (Vals AI) | Not measured | 54.7% | Only Mistral Large 4 measured |
| Tax Agent Bench (Vals AI) | 51.8% | 30.7% | Gemini 4 Carbon |
| Legal Research Bench (Vals AI) | 55.8% | 35.6% | Gemini 4 Carbon |
| Harvey's Legal Agent Benchmark (Vals AI) | 17.5% | 17.5% | Tie |
| Excel Modeling Benchmark (Vals AI) | Not measured | 56.2% | Only Mistral Large 4 measured |
| BioMysteryBench (Vals AI) | Not measured | 67.0% | Only Mistral Large 4 measured |
| APEX-Agents | 84.2% | 51.6% | Gemini 4 Carbon |
| Vending-Bench 2 | $16,484 | Not measured | Only Gemini 4 Carbon measured |
| Neither model has a result on the other 31 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index90
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| LiveBench reasoning | 90.6% | 85.2% | Gemini 4 Carbon |
| MysteryMechanism (Vals AI) | Not measured | 12.2% | Only Mistral Large 4 measured |
| HLE-Diamond (Artificial Analysis run) | 69.1% | 17.3% | Gemini 4 Carbon |
| CritPt v2, hard subset (Artificial Analysis run) | 69.0% | 27.0% | Gemini 4 Carbon |
| GPQA Diamond (Epoch AI run) | 96.5% | 88.1% | Gemini 4 Carbon |
| ARC-AGI | 98.5% | Not measured | Only Gemini 4 Carbon measured |
| ARC-AGI-2 | 94.2% | Not measured | Only Gemini 4 Carbon measured |
| SimpleBench | 92.6% | Not measured | Only Gemini 4 Carbon measured |
| Chess puzzles | 78.0% | 33.0% | Gemini 4 Carbon |
| Mystery game puzzles | 73.0% | 17.0% | Gemini 4 Carbon |
| EBR-bench | 87.6% | Not measured | Only Gemini 4 Carbon measured |
| LMArena Text: hard prompts | 1551 | 1434 | Gemini 4 Carbon |
| LMArena Text: expert | 1582 | 1446 | Gemini 4 Carbon |
| LMArena Text: hard prompts in English | 1545 | 1441 | Gemini 4 Carbon |
| LMArena Vision: diagrams | Not measured | 1272 | Only Mistral Large 4 measured |
| LMArena Vision: homework | Not measured | 1266 | Only Mistral Large 4 measured |
| Neither model has a result on the other 10 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 98.3% | Gemini 4 Carbon |
| FrontierMath, tiers 1 to 3 | 90.5% | 58.6% | Gemini 4 Carbon |
| FrontierMath, tier 4 | 95.1% | 26.8% | Gemini 4 Carbon |
| ProofBench | Not measured | 10.0% | Only Mistral Large 4 measured |
| LMArena Text: math | Not measured | 1456 | Only Mistral Large 4 measured |
| LMArena Text: mathematical work | 1538 | 1450 | Gemini 4 Carbon |
| MathArena | 85.4% | 21.1% | Gemini 4 Carbon |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| MedCode (Vals AI) | Not measured | 40.7% | Only Mistral Large 4 measured |
| AA-Omniscience | 53.5 | 5.1 | Gemini 4 Carbon |
| SimpleQA Verified | 77.7% | 20.0% | Gemini 4 Carbon |
| LMArena Text: science | 1535 | 1432 | Gemini 4 Carbon |
| LMArena Text: medicine and healthcare | 1525 | 1391 | Gemini 4 Carbon |
| LMArena Text: legal and government | 1542 | 1428 | Gemini 4 Carbon |
| LMArena Text: business and finance | 1520 | 1434 | Gemini 4 Carbon |
| FACTS Benchmark Suite | 78.4% | Not measured | Only Gemini 4 Carbon measured |
| PRBench Legal | 53.3% | Not measured | Only Gemini 4 Carbon measured |
| Neither model has a result on the other 12 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| Design Arena: websites | 1319 | 1258 | Gemini 4 Carbon |
| Design Arena: UI components | 1357 | 1278 | Gemini 4 Carbon |
| Design Arena: game dev | 1326 | 1204 | Gemini 4 Carbon |
| Design Arena: data viz | Not measured | 1273 | Only Mistral Large 4 measured |
| Design Arena: 3D | 1434 | 1238 | Gemini 4 Carbon |
| Design Arena: all code categories | 1348 | 1255 | Gemini 4 Carbon |
| Design Arena: SVG | Not measured | 1222 | Only Mistral Large 4 measured |
| Design Arena: ASCII art | Not measured | 1203 | Only Mistral Large 4 measured |
| LiveBench design | 80.2% | 76.3% | Gemini 4 Carbon |
GeneralLike-for-likePublic boards, shown as published, kept out of the index140
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | Leader |
|---|---|---|---|
| Vals Index | 70.5% | 56.5% | Gemini 4 Carbon |
| LiveBench data analysis | 86.5% | 76.6% | Gemini 4 Carbon |
| MedScribe (Vals AI) | Not measured | 80.4% | Only Mistral Large 4 measured |
| Long context reasoning (AA-LCR) | 79.3% | Not measured | Only Gemini 4 Carbon measured |
| GDPval-AA | 1733 | Not measured | Only Gemini 4 Carbon measured |
| DTBench | Not measured | 88.2% | Only Mistral Large 4 measured |
| LMCA | Not measured | 35.8% | Only Mistral Large 4 measured |
| Surface Evolver Bench | Not measured | 28.1% | Only Mistral Large 4 measured |
| Furniture assembly | 77.5% | 20.8% | Gemini 4 Carbon |
| LMArena Text | 1538 | 1411 | Gemini 4 Carbon |
| LMArena Text: instruction following | 1539 | 1409 | Gemini 4 Carbon |
| LMArena Text: creative writing | 1533 | 1343 | Gemini 4 Carbon |
| LMArena Vision | Not measured | 1244 | Only Mistral Large 4 measured |
| LMArena Text: English | 1530 | 1424 | Gemini 4 Carbon |
| LMArena Text: non-English | 1534 | 1397 | Gemini 4 Carbon |
| LMArena Text: Chinese | 1613 | 1457 | Gemini 4 Carbon |
| LMArena Text: French | Not measured | 1432 | Only Mistral Large 4 measured |
| LMArena Text: Spanish | Not measured | 1454 | Only Mistral Large 4 measured |
| LMArena Text: Russian | 1560 | 1414 | Gemini 4 Carbon |
| LMArena Text: multi-turn | 1569 | 1421 | Gemini 4 Carbon |
| LMArena Text: longer queries | 1549 | 1404 | Gemini 4 Carbon |
| LMArena Text: writing and language | 1542 | 1370 | Gemini 4 Carbon |
| LMArena Text: entertainment and media | 1527 | 1363 | Gemini 4 Carbon |
| LMArena Vision: English | Not measured | 1254 | Only Mistral Large 4 measured |
| LMArena Vision: reading text | Not measured | 1254 | Only Mistral Large 4 measured |
| LMArena Vision: creative writing | Not measured | 1186 | Only Mistral Large 4 measured |
| Neither model has a result on the other 23 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Gemini 4 Carbon | Mistral Large 4 | 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
Mistral Large 4 costs less
Repository review
50K input + 3K output tokens
Mistral Large 4 costs less
Agent loop
200K input + 10K output tokens
Mistral Large 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 | Google Gemini 4 Carbon | Mistral Large 4 |
|---|---|---|
| Maker | Mistral AI | |
| Weights | No public weights listed | Open weights |
| Released | November 10, 2026 | October 6, 2026 |
| Context window | 1.05M | 1.05M |
| Input $/Mtok | $2.00 | $0.70 |
| Output $/Mtok | $10.00 | $2.10 |
| Coding index | 86.6 | 60.9 |
| Independent measurements behind it | 4 | 5 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Gemini 4 Carbon better than Mistral Large 4 for coding?
Pick Gemini 4 Carbon for coding: it scores 86.6 on the Vibe Coding Index against 60.9 for Mistral Large 4, though on 4 benchmarks against 5. Pick Mistral Large 4 for output tokens 4.8x cheaper ($2.10 vs $10.00 per million).
Which is cheaper, Gemini 4 Carbon or Mistral Large 4?
Gemini 4 Carbon costs $2.00 per million input tokens and $10.00 per million output tokens; Mistral Large 4 costs $0.70 and $2.10. Cost per solved task is known for neither of them.
Which has the bigger context window, Gemini 4 Carbon or Mistral Large 4?
Neither: both take 1.05M tokens.
How many benchmarks is the Gemini 4 Carbon vs Mistral Large 4 comparison based on?
60 benchmarks have a result for both models as of November 14, 2026. Gemini 4 Carbon's coding index rests on 4 independent benchmarks and Mistral Large 4's on 5. Where only one model has a result, the page says so instead of naming a winner.