Claude Opus 5.5 vs Gemini 4 Carbon
Claude Opus 5.5 leads Gemini 4 Carbon on coding, 87.3 to 86.6 on the Vibe Coding Index, and Gemini 4 Carbon costs 2.0x less per output token ($10.00 vs $20.00 per million). Updated November 14, 2026; 73 benchmarks have a result for both.
Claude Opus 5.5 or Gemini 4 Carbon: which to pick
Claude Opus 5.5 and Gemini 4 Carbon are level for coding, 87.3 against 86.6 on the Vibe Coding Index, so price and context decide.
Pick Gemini 4 Carbon for output tokens 2.0x cheaper ($10.00 vs $20.00 per million) and a larger context window (1.05M against 1M tokens). Pick Claude Opus 5.5 for work measured by LMArena WebDev or Terminal-Bench 4 (Vals AI), where it leads.
- Shared results
- 73
- Claude Opus 5.5 only
- 65
- Gemini 4 Carbon only
- 0
- 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
No call
Level: 87.3 against 86.6, inside 2 points.
5 and 4 benchmarks behind the scores
Intelligence
AA Intelligence Index
Gemini 4 Carbon
Gemini 4 Carbon leads, 60.0 against 59.1.
Same board, both measured
Web design
Design Arena: websites
Claude Opus 5.5
Claude Opus 5.5 leads, 1326 against 1319.
Same board, both measured
General
Vals Index
Gemini 4 Carbon
Gemini 4 Carbon leads, 70.5% against 70.1%.
Same board, both measured
Long documents
Listed context window
Gemini 4 Carbon
Gemini 4 Carbon takes 1.05M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemini 4 Carbon
Claude Opus 5.5 $0.014, Gemini 4 Carbon $0.0070 per turn.
List prices
At a glance
- Coding index
- Claude Opus 5.5 87.3 from 5 benchmarks, Gemini 4 Carbon 86.6 from 4 benchmarks.
- Widest category
- Agentic work, 10.0 points apart on average over 10 shared benchmarks.
- Widest single gap
- Vending-Bench 2: Claude Opus 5.5 $9,235 against Gemini 4 Carbon $16,484, 42.3 points apart on a 0 to 100 scale.
- Output price
- Claude Opus 5.5 $20.00 against Gemini 4 Carbon $10.00 per million output tokens, 2.0x apart.
- Context window
- Claude Opus 5.5 takes 1M tokens, Gemini 4 Carbon 1.05M.
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 | Claude Opus 5.5 | Bars | Gemini 4 Carbon | Ties |
|---|---|---|---|---|
| Coding | 6 | 7 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 4 | 6 | ||
| Reasoning | 1 | 11 | 1 tie | |
| Math | 1 | 3 | 1 tie | |
| Knowledge | 0 | 8 | ||
| Web design | 6 | 0 | ||
| General | 3 | 13 | ||
| All shared boards | 21 | 50 | 2 ties |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index67
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| LiveBench coding | 87.1% | 91.0% | Gemini 4 Carbon |
| LMArena WebDev | 1815 | 1768 | Claude Opus 5.5 |
| Terminal-Bench 4 (Vals AI) | 65.2% | 63.1% | Claude Opus 5.5 |
| Vibe Code Bench (Vals AI) | 90.3% | 89.1% | Claude Opus 5.5 |
| Code migration (Vals AI) | 66.7% | Not measured | Only Claude Opus 5.5 measured |
| ProgramBench (Vals AI) | 18.5% | Not measured | Only Claude Opus 5.5 measured |
| IOI (Vals AI) | 95.1% | Not measured | Only Claude Opus 5.5 measured |
| Terminal-Bench 2.1 (Vals AI) | 87.6% | Not measured | Only Claude Opus 5.5 measured |
| SWE-bench Multimodal (Vals AI) | 70.2% | Not measured | Only Claude Opus 5.5 measured |
| Vibe Code Bench 1-100 (Vals AI) | 30.4% | Not measured | Only Claude Opus 5.5 measured |
| CyberBench (Vals AI) | 74.6% | Not measured | Only Claude Opus 5.5 measured |
| SciCode | 66.9% | Not measured | Only Claude Opus 5.5 measured |
| Terminal-Bench 4.0 (Artificial Analysis run) | 59.6% | 65.2% | Gemini 4 Carbon |
| FrontierCode 2.0 | 46.4% | 40.6% | Claude Opus 5.5 |
| FrontierCode | 54.6% | Not measured | Only Claude Opus 5.5 measured |
| FrontierSWE | 62.3% | 61.2% | Claude Opus 5.5 |
| MirrorCode | 95.6% | Not measured | Only Claude Opus 5.5 measured |
| ALE-Bench | 2147.4 | 2374.7 | Gemini 4 Carbon |
| LMArena Text: coding | 1543 | 1564 | Gemini 4 Carbon |
| LMArena Text: software and IT | 1533 | 1552 | Gemini 4 Carbon |
| LMArena WebDev: React | 1868 | 1792 | Claude Opus 5.5 |
| LMArena WebDev: HTML | 1727 | 1770 | Gemini 4 Carbon |
| LMArena WebDev: from an image | 1752 | Not measured | Only Claude Opus 5.5 measured |
| SciCode | 18.5% | Not measured | Only Claude Opus 5.5 measured |
| GameBench 2 | 94.8 | Not measured | Only Claude Opus 5.5 measured |
| Terminal-Bench 4.0 | 64.8% | Not measured | Only Claude Opus 5.5 measured |
| CursorBench 4.0 | 57.8% | 60.4% | Gemini 4 Carbon |
| ProgramBench | 18.5% | Not measured | Only Claude Opus 5.5 measured |
| Neither model has a result on the other 24 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| AA Intelligence Index | 59.1 | 60.0 | Gemini 4 Carbon |
| Epoch Capabilities Index | 168.5 | 169.8 | Gemini 4 Carbon |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index46
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| Design Arena: agents building web apps | 1321 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: agents building full-stack apps | 1390 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: agents building mobile apps | 1294 | 1284 | Claude Opus 5.5 |
| Design Arena: agents building Android apps | 1392 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: agents building games | 1456 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: HTML slides | 1471 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: PowerPoint slides | 1391 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: spreadsheets | 1328 | 1224 | Claude Opus 5.5 |
| LiveBench research | 88.9% | 92.5% | Gemini 4 Carbon |
| LiveBench browser | 88.4% | 80.0% | Claude Opus 5.5 |
| LiveBench long agentic | 72.8% | 72.7% | Claude Opus 5.5 |
| SREBench (Vals AI) | 33.6% | Not measured | Only Claude Opus 5.5 measured |
| Finance Agent (Vals AI) | 58.6% | Not measured | Only Claude Opus 5.5 measured |
| Tax Agent Bench (Vals AI) | 44.5% | 51.8% | Gemini 4 Carbon |
| Legal Research Bench (Vals AI) | 50.5% | 55.8% | Gemini 4 Carbon |
| Harvey's Legal Agent Benchmark (Vals AI) | 9.2% | 17.5% | Gemini 4 Carbon |
| Excel Modeling Benchmark (Vals AI) | 75.9% | Not measured | Only Claude Opus 5.5 measured |
| BioMysteryBench (Vals AI) | 79.3% | Not measured | Only Claude Opus 5.5 measured |
| Terminal-Bench-Science (Vals AI) | 47.1% | Not measured | Only Claude Opus 5.5 measured |
| Vals Science Index | 58.7% | Not measured | Only Claude Opus 5.5 measured |
| Time Horizon Index (Vals AI) | 91.3% | Not measured | Only Claude Opus 5.5 measured |
| CUA-bench (Vals AI) | 14.0% | Not measured | Only Claude Opus 5.5 measured |
| Robo-bench (Vals AI) | 14.6% | Not measured | Only Claude Opus 5.5 measured |
| APEX-Agents | 73.5% | 84.2% | Gemini 4 Carbon |
| Remote Labor Index | 21.3% | Not measured | Only Claude Opus 5.5 measured |
| PostTrainBench | 43.8% | Not measured | Only Claude Opus 5.5 measured |
| SpireClimb | 5.8% | Not measured | Only Claude Opus 5.5 measured |
| Vending-Bench 2 | $9,235 | $16,484 | Gemini 4 Carbon |
| LMArena Search | 1213 | Not measured | Only Claude Opus 5.5 measured |
| MCP Atlas | 85.4% | Not measured | Only Claude Opus 5.5 measured |
| Neither model has a result on the other 14 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index111
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| LiveBench reasoning | 92.2% | 90.6% | Claude Opus 5.5 |
| MysteryMechanism (Vals AI) | 49.5% | Not measured | Only Claude Opus 5.5 measured |
| Humanity's Last Exam (Artificial Analysis run) | 61.4% | Not measured | Only Claude Opus 5.5 measured |
| HLE-Diamond (Artificial Analysis run) | 67.9% | 69.1% | Gemini 4 Carbon |
| CritPt v2, hard subset (Artificial Analysis run) | 68.0% | 69.0% | Gemini 4 Carbon |
| GPQA Diamond (Epoch AI run) | 90.6% | 96.5% | Gemini 4 Carbon |
| ARC-AGI | 98.5% | 98.5% | Tie |
| ARC-AGI-2 | 93.3% | 94.2% | Gemini 4 Carbon |
| SimpleBench | 88.4% | 92.6% | Gemini 4 Carbon |
| Chess puzzles | 67.0% | 78.0% | Gemini 4 Carbon |
| Mystery game puzzles | 71.0% | 73.0% | Gemini 4 Carbon |
| EBR-bench | 71.4% | 87.6% | Gemini 4 Carbon |
| LMArena Text: hard prompts | 1531 | 1551 | Gemini 4 Carbon |
| LMArena Text: expert | 1555 | 1582 | Gemini 4 Carbon |
| LMArena Text: hard prompts in English | 1536 | 1545 | Gemini 4 Carbon |
| LMArena Vision: diagrams | 1354 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Vision: homework | 1357 | Not measured | Only Claude Opus 5.5 measured |
| Humanity's Last Exam (Scale run) | 55.0% | Not measured | Only Claude Opus 5.5 measured |
| GoBench | 2274 | Not measured | Only Claude Opus 5.5 measured |
| Neither model has a result on the other 7 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index13
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| OTIS Mock AIME | 100.0% | 100.0% | Tie |
| FrontierMath, tiers 1 to 3 | 91.2% | 90.5% | Claude Opus 5.5 |
| FrontierMath, tier 4 | 95.0% | 95.1% | Gemini 4 Carbon |
| ProofBench | 100.0% | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: math | 1520 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: mathematical work | 1532 | 1538 | Gemini 4 Carbon |
| MathArena | 75.3% | 85.4% | 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 index08
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| MedCode (Vals AI) | 49.8% | Not measured | Only Claude Opus 5.5 measured |
| AA-Omniscience | 48.1 | 53.5 | Gemini 4 Carbon |
| SimpleQA Verified | 72.2% | 77.7% | Gemini 4 Carbon |
| LMArena Text: science | 1525 | 1535 | Gemini 4 Carbon |
| LMArena Text: medicine and healthcare | 1511 | 1525 | Gemini 4 Carbon |
| LMArena Text: legal and government | 1519 | 1542 | Gemini 4 Carbon |
| LMArena Text: business and finance | 1494 | 1520 | Gemini 4 Carbon |
| FACTS Benchmark Suite | 66.5% | 78.4% | Gemini 4 Carbon |
| PRBench Legal | 47.6% | 53.3% | Gemini 4 Carbon |
| 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 | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| Design Arena: websites | 1326 | 1319 | Claude Opus 5.5 |
| Design Arena: UI components | 1377 | 1357 | Claude Opus 5.5 |
| Design Arena: game dev | 1427 | 1326 | Claude Opus 5.5 |
| Design Arena: data viz | 1387 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: 3D | 1460 | 1434 | Claude Opus 5.5 |
| Design Arena: all code categories | 1384 | 1348 | Claude Opus 5.5 |
| Design Arena: SVG | 1320 | Not measured | Only Claude Opus 5.5 measured |
| Design Arena: ASCII art | 1354 | Not measured | Only Claude Opus 5.5 measured |
| LiveBench design | 83.3% | 80.2% | Claude Opus 5.5 |
GeneralLike-for-likePublic boards, shown as published, kept out of the index313
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | Leader |
|---|---|---|---|
| Vals Index | 70.1% | 70.5% | Gemini 4 Carbon |
| LiveBench data analysis | 80.3% | 86.5% | Gemini 4 Carbon |
| Public Benefits Bench (Vals AI) | 70.6% | Not measured | Only Claude Opus 5.5 measured |
| SAGE (Vals AI) | 91.3% | Not measured | Only Claude Opus 5.5 measured |
| MedScribe (Vals AI) | 91.4% | Not measured | Only Claude Opus 5.5 measured |
| Long context reasoning (AA-LCR) | 84.7% | 79.3% | Claude Opus 5.5 |
| GDPval-AA | 1866 | 1733 | Claude Opus 5.5 |
| DTBench | 98.9% | Not measured | Only Claude Opus 5.5 measured |
| LMCA | 68.2% | Not measured | Only Claude Opus 5.5 measured |
| Furniture assembly | 83.3% | 77.5% | Claude Opus 5.5 |
| LMArena Text | 1514 | 1538 | Gemini 4 Carbon |
| LMArena Text: instruction following | 1535 | 1539 | Gemini 4 Carbon |
| LMArena Text: creative writing | 1532 | 1533 | Gemini 4 Carbon |
| LMArena Vision | 1324 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: English | 1521 | 1530 | Gemini 4 Carbon |
| LMArena Text: non-English | 1503 | 1534 | Gemini 4 Carbon |
| LMArena Text: Chinese | 1573 | 1613 | Gemini 4 Carbon |
| LMArena Text: German | 1513 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: Japanese | 1527 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: Korean | 1505 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: French | 1498 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: Spanish | 1510 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: Russian | 1517 | 1560 | Gemini 4 Carbon |
| LMArena Text: Polish | 1511 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Text: multi-turn | 1502 | 1569 | Gemini 4 Carbon |
| LMArena Text: longer queries | 1528 | 1549 | Gemini 4 Carbon |
| LMArena Text: writing and language | 1528 | 1542 | Gemini 4 Carbon |
| LMArena Text: entertainment and media | 1508 | 1527 | Gemini 4 Carbon |
| LMArena Vision: English | 1321 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Vision: Chinese | 1380 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Vision: reading text | 1332 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Vision: humor | 1267 | Not measured | Only Claude Opus 5.5 measured |
| LMArena Vision: creative writing | 1317 | Not measured | Only Claude Opus 5.5 measured |
| EQ-Bench 4 | 1340 | Not measured | Only Claude Opus 5.5 measured |
| Neither model has a result on the other 15 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Claude Opus 5.5 | Gemini 4 Carbon | 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
Gemini 4 Carbon costs less
Repository review
50K input + 3K output tokens
Gemini 4 Carbon costs less
Agent loop
200K input + 10K output tokens
Gemini 4 Carbon 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 | Anthropic Claude Opus 5.5 | Google Gemini 4 Carbon |
|---|---|---|
| Maker | Anthropic | |
| Weights | No public weights listed | No public weights listed |
| Released | September 22, 2026 | November 10, 2026 |
| Context window | 1M | 1.05M |
| Input $/Mtok | $4.00 | $2.00 |
| Output $/Mtok | $20.00 | $10.00 |
| Coding index | 87.3 | 86.6 |
| Independent measurements behind it | 5 | 4 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Opus 5.5 better than Gemini 4 Carbon for coding?
Claude Opus 5.5 and Gemini 4 Carbon are level for coding, 87.3 against 86.6 on the Vibe Coding Index, so price and context decide. Pick Gemini 4 Carbon for output tokens 2.0x cheaper ($10.00 vs $20.00 per million) and a larger context window (1.05M against 1M tokens). Pick Claude Opus 5.5 for work measured by LMArena WebDev or Terminal-Bench 4 (Vals AI), where it leads.
Which is cheaper, Claude Opus 5.5 or Gemini 4 Carbon?
Claude Opus 5.5 costs $4.00 per million input tokens and $20.00 per million output tokens; Gemini 4 Carbon costs $2.00 and $10.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Claude Opus 5.5 or Gemini 4 Carbon?
Gemini 4 Carbon, with 1.05M tokens against 1M.
How many benchmarks is the Claude Opus 5.5 vs Gemini 4 Carbon comparison based on?
73 benchmarks have a result for both models as of November 14, 2026. Claude Opus 5.5's coding index rests on 5 independent benchmarks and Gemini 4 Carbon's on 4. Where only one model has a result, the page says so instead of naming a winner.