Claude Haiku 5.5 vs Claude Sonnet 5.5
Updated October 21, 2026. 58 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Pick Claude Sonnet 5.5 for coding: it scores 88.9 on the Vibe Coding Index against 75.9 for Claude Haiku 5.5.
Pick Claude Haiku 5.5 when output tokens cost 20.0x less, or when the benchmark that matters is CyberBench (Vals AI), where it leads.
- Shared results
- 58
- Claude Haiku 5.5 only
- 4
- Claude Sonnet 5.5 only
- 37
- Like-for-like categories
- 8 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, four public coding benchmarks
Claude Sonnet 5.5
Claude Sonnet 5.5 scores 88.9 against 75.9.
2 benchmarks behind each score
Intelligence
AA Intelligence Index
Claude Sonnet 5.5
Claude Sonnet 5.5 leads, 56.0 against 43.4.
Same board, both measured
Web design
Design Arena: websites
Claude Sonnet 5.5
Claude Sonnet 5.5 leads, 1326 against 1263.
Same board, both measured
General
Vals Index
Claude Sonnet 5.5
Claude Sonnet 5.5 leads, 70.8% against 58.7%.
Same board, both measured
Long documents
Listed context window
No call
Both take 1M tokens.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Claude Haiku 5.5
Claude Haiku 5.5 $0.0003, Claude Sonnet 5.5 $0.0070 per turn.
List prices
At a glance
- Coding index
- Claude Haiku 5.5 75.9 from 2 benchmarks, Claude Sonnet 5.5 88.9 from 2 benchmarks.
- Widest category
- Agentic work, 16.3 points apart on average over 14 shared benchmarks.
- Widest single gap
- Vending-Bench 2: Claude Haiku 5.5 $3,602 against Claude Sonnet 5.5 $11,165, 47.3 points apart on a 0 to 100 scale.
- Output price
- Claude Haiku 5.5 $0.50 against Claude Sonnet 5.5 $10.00 per million output tokens, 20.0x apart.
- Context window
- Both take 1M 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 | Claude Haiku 5.5 | Bars | Claude Sonnet 5.5 | Ties |
|---|---|---|---|---|
| Coding | 1 | 13 | ||
| Intelligence | 0 | 2 | ||
| Agentic work | 0 | 14 | ||
| Reasoning | 0 | 6 | ||
| Math | 0 | 4 | ||
| Knowledge | 0 | 2 | ||
| Web design | 0 | 7 | ||
| General | 0 | 8 | 1 tie | |
| All shared boards | 1 | 56 | 1 tie |
CodingLike-for-likeFour of these boards feed the Vibe Coding Index113
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| LiveBench coding | 91.4% | 94.7% | Claude Sonnet 5.5 |
| LMArena WebDev | 1586 | 1768 | Claude Sonnet 5.5 |
| Terminal-Bench 4 (Vals AI) | 36.4% | 64.1% | Claude Sonnet 5.5 |
| Vibe Code Bench (Vals AI) | 90.4% | 92.4% | Claude Sonnet 5.5 |
| Code migration (Vals AI) | 44.6% | 69.8% | Claude Sonnet 5.5 |
| ProgramBench (Vals AI) | 2.5% | 6.5% | Claude Sonnet 5.5 |
| IOI (Vals AI) | 47.3% | 83.1% | Claude Sonnet 5.5 |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 83.2% | Only Claude Sonnet 5.5 measured |
| SWE-bench Multimodal (Vals AI) | 44.0% | 58.3% | Claude Sonnet 5.5 |
| Vibe Code Bench 1-100 (Vals AI) | 20.4% | Not measured | Only Claude Haiku 5.5 measured |
| CyberBench (Vals AI) | 75.5% | 59.6% | Claude Haiku 5.5 |
| SciCode | 55.0% | 61.0% | Claude Sonnet 5.5 |
| FrontierCode | 46.4% | 52.1% | Claude Sonnet 5.5 |
| FrontierSWE | 43.8% | 61.9% | Claude Sonnet 5.5 |
| ALE-Bench | Not measured | 1819.1 | Only Claude Sonnet 5.5 measured |
| LMArena Text: coding | Not measured | 1513 | Only Claude Sonnet 5.5 measured |
| LMArena Text: software and IT | Not measured | 1504 | Only Claude Sonnet 5.5 measured |
| LMArena WebDev: React | 1605 | 1817 | Claude Sonnet 5.5 |
| LMArena WebDev: HTML | 1569 | 1741 | Claude Sonnet 5.5 |
| LMArena WebDev: from an image | Not measured | 1737 | Only Claude Sonnet 5.5 measured |
| Neither model has a result on the other 18 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| AA Intelligence Index | 43.4 | 56.0 | Claude Sonnet 5.5 |
| Epoch Capabilities Index | 159.3 | 166.2 | Claude Sonnet 5.5 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index014
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| Design Arena: agents building web apps | Not measured | 1303 | Only Claude Sonnet 5.5 measured |
| Design Arena: agents building mobile apps | Not measured | 1265 | Only Claude Sonnet 5.5 measured |
| Design Arena: spreadsheets | Not measured | 1245 | Only Claude Sonnet 5.5 measured |
| LiveBench research | 81.9% | 83.6% | Claude Sonnet 5.5 |
| LiveBench browser | 80.0% | 84.0% | Claude Sonnet 5.5 |
| LiveBench long agentic | 57.6% | 78.8% | Claude Sonnet 5.5 |
| SREBench (Vals AI) | 5.7% | 30.1% | Claude Sonnet 5.5 |
| Finance Agent (Vals AI) | 54.1% | 58.1% | Claude Sonnet 5.5 |
| Tax Agent Bench (Vals AI) | 31.9% | 42.0% | Claude Sonnet 5.5 |
| Legal Research Bench (Vals AI) | 43.3% | 48.1% | Claude Sonnet 5.5 |
| Harvey's Legal Agent Benchmark (Vals AI) | 4.2% | 10.4% | Claude Sonnet 5.5 |
| Excel Modeling Benchmark (Vals AI) | 59.0% | 75.7% | Claude Sonnet 5.5 |
| BioMysteryBench (Vals AI) | 70.4% | 81.1% | Claude Sonnet 5.5 |
| Terminal-Bench-Science (Vals AI) | 10.0% | 47.1% | Claude Sonnet 5.5 |
| Vals Science Index | 38.5% | 59.1% | Claude Sonnet 5.5 |
| APEX-Agents | 56.1% | 75.5% | Claude Sonnet 5.5 |
| SpireClimb | Not measured | 1.3% | Only Claude Sonnet 5.5 measured |
| Vending-Bench 2 | $3,602 | $11,165 | Claude Sonnet 5.5 |
| Neither model has a result on the other 21 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index06
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| LiveBench reasoning | 82.4% | 91.6% | Claude Sonnet 5.5 |
| MysteryMechanism (Vals AI) | 35.1% | 49.1% | Claude Sonnet 5.5 |
| Humanity's Last Exam (Artificial Analysis run) | 44.4% | 55.0% | Claude Sonnet 5.5 |
| GPQA Diamond (Epoch AI run) | 89.6% | 95.6% | Claude Sonnet 5.5 |
| ARC-AGI | 89.0% | Not measured | Only Claude Haiku 5.5 measured |
| ARC-AGI-2 | 61.3% | Not measured | Only Claude Haiku 5.5 measured |
| Chess puzzles | 55.0% | 59.0% | Claude Sonnet 5.5 |
| Mystery game puzzles | 30.0% | 65.0% | Claude Sonnet 5.5 |
| EBR-bench | 43.8% | Not measured | Only Claude Haiku 5.5 measured |
| LMArena Text: hard prompts | Not measured | 1494 | Only Claude Sonnet 5.5 measured |
| LMArena Text: expert | Not measured | 1531 | Only Claude Sonnet 5.5 measured |
| LMArena Text: hard prompts in English | Not measured | 1502 | Only Claude Sonnet 5.5 measured |
| LMArena Vision: diagrams | Not measured | 1329 | Only Claude Sonnet 5.5 measured |
| LMArena Vision: homework | Not measured | 1331 | Only Claude Sonnet 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 index04
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| OTIS Mock AIME | 98.9% | 100.0% | Claude Sonnet 5.5 |
| FrontierMath, tiers 1 to 3 | 75.1% | 88.8% | Claude Sonnet 5.5 |
| FrontierMath, tier 4 | 46.3% | 80.5% | Claude Sonnet 5.5 |
| ProofBench | 89.0% | 100.0% | Claude Sonnet 5.5 |
| LMArena Text: math | Not measured | 1497 | Only Claude Sonnet 5.5 measured |
| LMArena Text: mathematical work | Not measured | 1510 | Only Claude Sonnet 5.5 measured |
| Neither model has a result on the other 7 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index02
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| MedCode (Vals AI) | 51.5% | 52.9% | Claude Sonnet 5.5 |
| SimpleQA Verified | 23.8% | 46.5% | Claude Sonnet 5.5 |
| LMArena Text: science | Not measured | 1483 | Only Claude Sonnet 5.5 measured |
| LMArena Text: medicine and healthcare | Not measured | 1464 | Only Claude Sonnet 5.5 measured |
| LMArena Text: legal and government | Not measured | 1483 | Only Claude Sonnet 5.5 measured |
| LMArena Text: business and finance | Not measured | 1468 | Only Claude Sonnet 5.5 measured |
| Neither model has a result on the other 11 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index07
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| Design Arena: websites | 1263 | 1326 | Claude Sonnet 5.5 |
| Design Arena: UI components | 1295 | 1367 | Claude Sonnet 5.5 |
| Design Arena: game dev | 1315 | 1469 | Claude Sonnet 5.5 |
| Design Arena: data viz | 1292 | 1389 | Claude Sonnet 5.5 |
| Design Arena: 3D | 1343 | 1458 | Claude Sonnet 5.5 |
| Design Arena: all code categories | 1279 | 1364 | Claude Sonnet 5.5 |
| LiveBench design | 83.1% | 85.0% | Claude Sonnet 5.5 |
| Neither model has a result on the other 2 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index08
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | Leader |
|---|---|---|---|
| Vals Index | 58.7% | 70.8% | Claude Sonnet 5.5 |
| LiveBench data analysis | 53.1% | 59.5% | Claude Sonnet 5.5 |
| Public Benefits Bench (Vals AI) | 63.2% | 67.2% | Claude Sonnet 5.5 |
| SAGE (Vals AI) | 45.8% | 51.8% | Claude Sonnet 5.5 |
| MedScribe (Vals AI) | 89.5% | 91.1% | Claude Sonnet 5.5 |
| Long context reasoning (AA-LCR) | 82.7% | 82.7% | Tie |
| DTBench | 95.0% | 97.3% | Claude Sonnet 5.5 |
| LMCA | 51.1% | 68.3% | Claude Sonnet 5.5 |
| Furniture assembly | 47.5% | 75.0% | Claude Sonnet 5.5 |
| LMArena Text | Not measured | 1473 | Only Claude Sonnet 5.5 measured |
| LMArena Text: instruction following | Not measured | 1492 | Only Claude Sonnet 5.5 measured |
| LMArena Text: creative writing | Not measured | 1467 | Only Claude Sonnet 5.5 measured |
| LMArena Vision | Not measured | 1287 | Only Claude Sonnet 5.5 measured |
| LMArena Text: English | Not measured | 1484 | Only Claude Sonnet 5.5 measured |
| LMArena Text: non-English | Not measured | 1456 | Only Claude Sonnet 5.5 measured |
| LMArena Text: Chinese | Not measured | 1535 | Only Claude Sonnet 5.5 measured |
| LMArena Text: French | Not measured | 1499 | Only Claude Sonnet 5.5 measured |
| LMArena Text: Spanish | Not measured | 1451 | Only Claude Sonnet 5.5 measured |
| LMArena Text: Russian | Not measured | 1454 | Only Claude Sonnet 5.5 measured |
| LMArena Text: multi-turn | Not measured | 1465 | Only Claude Sonnet 5.5 measured |
| LMArena Text: longer queries | Not measured | 1495 | Only Claude Sonnet 5.5 measured |
| LMArena Text: writing and language | Not measured | 1474 | Only Claude Sonnet 5.5 measured |
| LMArena Text: entertainment and media | Not measured | 1461 | Only Claude Sonnet 5.5 measured |
| LMArena Vision: English | Not measured | 1289 | Only Claude Sonnet 5.5 measured |
| LMArena Vision: reading text | Not measured | 1297 | Only Claude Sonnet 5.5 measured |
| LMArena Vision: creative writing | Not measured | 1242 | Only Claude Sonnet 5.5 measured |
| Neither model has a result on the other 20 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Claude Haiku 5.5 | Claude Sonnet 5.5 | 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
Claude Haiku 5.5 costs less
Repository review
50K input + 3K output tokens
Claude Haiku 5.5 costs less
Agent loop
200K input + 10K output tokens
Claude Haiku 5.5 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 Haiku 5.5 | Anthropic Claude Sonnet 5.5 |
|---|---|---|
| Maker | Anthropic | Anthropic |
| Weights | No public weights listed | No public weights listed |
| Released | October 7, 2026 | September 28, 2026 |
| Context window | 1M | 1M |
| Input $/Mtok | $0.10 | $2.00 |
| Output $/Mtok | $0.50 | $10.00 |
| Coding index | 75.9 | 88.9 |
| Independent measurements behind it | 2 | 2 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Haiku 5.5 better than Claude Sonnet 5.5 for coding?
Pick Claude Sonnet 5.5 for coding: it scores 88.9 on the Vibe Coding Index against 75.9 for Claude Haiku 5.5. Pick Claude Haiku 5.5 when output tokens cost 20.0x less, or when the benchmark that matters is CyberBench (Vals AI), where it leads.
Which is cheaper, Claude Haiku 5.5 or Claude Sonnet 5.5?
Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens; Claude Sonnet 5.5 costs $2.00 and $10.00. Cost per solved task is known for neither of them.
Which has the bigger context window, Claude Haiku 5.5 or Claude Sonnet 5.5?
Neither: both take 1M tokens.
How many benchmarks is the Claude Haiku 5.5 vs Claude Sonnet 5.5 comparison based on?
58 benchmarks have a result for both models as of October 21, 2026. Claude Haiku 5.5's coding index rests on 2 independent benchmarks and Claude Sonnet 5.5's on 2. Where only one model has a result, the page says so instead of naming a winner.