Claude Haiku 5.5 vs Grok 4.20
Claude Haiku 5.5 leads Grok 4.20 on coding, 70.7 to 33.3 on the Vibe Coding Index, and Claude Haiku 5.5 costs 5.0x less per output token ($0.50 vs $2.50 per million). Updated October 22, 2026; 24 benchmarks have a result for both.
Claude Haiku 5.5 or Grok 4.20: which to pick
Pick Claude Haiku 5.5 for coding: it scores 70.7 on the Vibe Coding Index against 33.3 for Grok 4.20.
Pick Grok 4.20 for a larger context window (2M against 1M tokens) and work measured by Vending-Bench 2 or ARC-AGI, where it leads.
- Shared results
- 24
- Claude Haiku 5.5 only
- 39
- Grok 4.20 only
- 31
- Like-for-like categories
- 7 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
Claude Haiku 5.5
Claude Haiku 5.5 scores 70.7 against 33.3.
5 and 2 benchmarks behind the scores
Intelligence
AA Intelligence Index
Claude Haiku 5.5
Claude Haiku 5.5 leads, 43.4 against 25.7.
Same board, both measured
Reasoning
GPQA Diamond
No call
Only Grok 4.20 has a GPQA Diamond result, 88.6%.
Directional only
Math
AIME
No call
Only Grok 4.20 has a AIME result, 96.5%.
Directional only
Knowledge
MMLU Pro
No call
Only Grok 4.20 has a MMLU Pro result, 86.3%.
Directional only
Web design
Design Arena: websites
Claude Haiku 5.5
Claude Haiku 5.5 leads, 1262 against 1229.
Same board, both measured
General
Vals Index
No call
Only Claude Haiku 5.5 has a Vals Index result, 58.7%.
Directional only
Long documents
Listed context window
Grok 4.20
Grok 4.20 takes 2M tokens against 1M.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Claude Haiku 5.5
Claude Haiku 5.5 $0.0003, Grok 4.20 $0.0025 per turn.
List prices
Cost per solved task
One SWE-bench Verified task, solved
No call
Only Grok 4.20 has the number, $0.72.
Directional only
At a glance
- Coding index
- Claude Haiku 5.5 70.7 from 5 benchmarks, Grok 4.20 33.3 from 2 benchmarks.
- Widest category
- Coding, 65.3 points apart on average over 2 shared benchmarks.
- Widest single gap
- Vibe Code Bench (Vals AI): Claude Haiku 5.5 90.4% against Grok 4.20 4.1%, 86.4 points apart on a 0 to 100 scale.
- Output price
- Claude Haiku 5.5 $0.50 against Grok 4.20 $2.50 per million output tokens, 5.0x apart.
- Context window
- Claude Haiku 5.5 takes 1M tokens, Grok 4.20 2M.
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 | Grok 4.20 | Ties |
|---|---|---|---|---|
| Coding | 2 | 0 | ||
| Intelligence | 2 | 0 | ||
| Agentic work | 4 | 1 | ||
| Reasoning | 1 | 2 | ||
| Knowledge | 1 | 0 | ||
| Web design | 6 | 0 | ||
| General | 5 | 0 | ||
| All shared boards | 21 | 3 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index20
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| LiveBench coding | 91.4% | Not measured | Only Claude Haiku 5.5 measured |
| LMArena WebDev | 1586 | Not measured | Only Claude Haiku 5.5 measured |
| LiveCodeBench (Vals AI) | Not measured | 84.3% | Only Grok 4.20 measured |
| SWE-bench Verified (Vals AI) | Not measured | 72.2% | Only Grok 4.20 measured |
| Terminal-Bench 4 (Vals AI) | 36.4% | Not measured | Only Claude Haiku 5.5 measured |
| Vibe Code Bench (Vals AI) | 90.4% | 4.1% | Claude Haiku 5.5 |
| Code migration (Vals AI) | 44.6% | 0.3% | Claude Haiku 5.5 |
| ProgramBench (Vals AI) | 2.5% | Not measured | Only Claude Haiku 5.5 measured |
| IOI (Vals AI) | 47.3% | Not measured | Only Claude Haiku 5.5 measured |
| Terminal-Bench 2.1 (Vals AI) | Not measured | 44.2% | Only Grok 4.20 measured |
| SWE-bench Multimodal (Vals AI) | 44.0% | Not measured | Only Claude Haiku 5.5 measured |
| Vibe Code Bench 1-100 (Vals AI) | 20.4% | Not measured | Only Claude Haiku 5.5 measured |
| CyberBench (Vals AI) | 75.5% | Not measured | Only Claude Haiku 5.5 measured |
| SciCode | 55.0% | Not measured | Only Claude Haiku 5.5 measured |
| Terminal-Bench Hard (Artificial Analysis run) | Not measured | 37.9% | Only Grok 4.20 measured |
| Terminal-Bench | Not measured | 57.3% | Only Grok 4.20 measured |
| FrontierCode | 46.4% | Not measured | Only Claude Haiku 5.5 measured |
| FrontierSWE | 43.8% | Not measured | Only Claude Haiku 5.5 measured |
| WeirdML | Not measured | 52.3% | Only Grok 4.20 measured |
| LMArena WebDev: React | 1605 | Not measured | Only Claude Haiku 5.5 measured |
| LMArena WebDev: HTML | 1569 | Not measured | Only Claude Haiku 5.5 measured |
| SciCode | 9.2% | Not measured | Only Claude Haiku 5.5 measured |
| Neither model has a result on the other 17 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index20
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| AA Intelligence Index | 43.4 | 25.7 | Claude Haiku 5.5 |
| Epoch Capabilities Index | 159.3 | 152.1 | Claude Haiku 5.5 |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index41
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| Design Arena: agents building web apps | Not measured | 1144 | Only Grok 4.20 measured |
| Design Arena: agents building full-stack apps | Not measured | 1054 | Only Grok 4.20 measured |
| Design Arena: agents building mobile apps | Not measured | 1090 | Only Grok 4.20 measured |
| Design Arena: agents building Android apps | Not measured | 1045 | Only Grok 4.20 measured |
| Design Arena: agents building Godot games | Not measured | 1089 | Only Grok 4.20 measured |
| Design Arena: HTML slides | Not measured | 1180 | Only Grok 4.20 measured |
| LiveBench research | 81.9% | Not measured | Only Claude Haiku 5.5 measured |
| LiveBench browser | 80.0% | Not measured | Only Claude Haiku 5.5 measured |
| LiveBench long agentic | 57.6% | Not measured | Only Claude Haiku 5.5 measured |
| SREBench (Vals AI) | 5.7% | Not measured | Only Claude Haiku 5.5 measured |
| Finance Agent (Vals AI) | 54.1% | 28.5% | Claude Haiku 5.5 |
| Tax Agent Bench (Vals AI) | 31.9% | 7.0% | Claude Haiku 5.5 |
| Legal Research Bench (Vals AI) | 43.3% | 13.9% | Claude Haiku 5.5 |
| Harvey's Legal Agent Benchmark (Vals AI) | 4.2% | Not measured | Only Claude Haiku 5.5 measured |
| Excel Modeling Benchmark (Vals AI) | 59.0% | 11.7% | Claude Haiku 5.5 |
| BioMysteryBench (Vals AI) | 70.4% | Not measured | Only Claude Haiku 5.5 measured |
| Terminal-Bench-Science (Vals AI) | 10.0% | Not measured | Only Claude Haiku 5.5 measured |
| Vals Science Index | 38.5% | Not measured | Only Claude Haiku 5.5 measured |
| τ²-Bench Telecom | Not measured | 93.0% | Only Grok 4.20 measured |
| APEX-Agents | 56.1% | Not measured | Only Claude Haiku 5.5 measured |
| Vending-Bench 2 | $3,602 | $4,663 | Grok 4.20 |
| Neither model has a result on the other 20 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index12
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| GPQA Diamond | Not measured | 88.6% | Only Grok 4.20 measured |
| LiveBench reasoning | 82.4% | Not measured | Only Claude Haiku 5.5 measured |
| MysteryMechanism (Vals AI) | 35.1% | Not measured | Only Claude Haiku 5.5 measured |
| MMMU Pro (Vals AI) | Not measured | 83.5% | Only Grok 4.20 measured |
| GPQA Diamond (Artificial Analysis run) | Not measured | 91.1% | Only Grok 4.20 measured |
| Humanity's Last Exam (Artificial Analysis run) | 44.4% | 34.5% | Claude Haiku 5.5 |
| GPQA Diamond (Epoch AI run) | 89.6% | Not measured | Only Claude Haiku 5.5 measured |
| ARC-AGI | 89.0% | 89.5% | Grok 4.20 |
| ARC-AGI-2 | 61.3% | 65.1% | Grok 4.20 |
| Chess puzzles | 55.0% | Not measured | Only Claude Haiku 5.5 measured |
| Mystery game puzzles | 30.0% | Not measured | Only Claude Haiku 5.5 measured |
| EBR-bench | 43.8% | Not measured | Only Claude Haiku 5.5 measured |
| Kagi LLM Benchmark | Not measured | 75.0% | Only Grok 4.20 measured |
| Neither model has a result on the other 9 boards here | |||
MathDirectional onlyPublic boards, shown as published, kept out of the index5 boards, none shared
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| AIME | Not measured | 96.5% | Only Grok 4.20 measured |
| OTIS Mock AIME | 98.9% | Not measured | Only Claude Haiku 5.5 measured |
| FrontierMath, tiers 1 to 3 | 75.1% | Not measured | Only Claude Haiku 5.5 measured |
| FrontierMath, tier 4 | 46.3% | Not measured | Only Claude Haiku 5.5 measured |
| ProofBench | 89.0% | Not measured | Only Claude Haiku 5.5 measured |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| MMLU Pro | Not measured | 86.3% | Only Grok 4.20 measured |
| LegalBench (Vals AI) | Not measured | 77.7% | Only Grok 4.20 measured |
| CaseLaw (Vals AI) | Not measured | 54.4% | Only Grok 4.20 measured |
| CorpFin (Vals AI) | Not measured | 63.7% | Only Grok 4.20 measured |
| TaxEval (Vals AI) | Not measured | 74.1% | Only Grok 4.20 measured |
| MortgageTax (Vals AI) | Not measured | 45.4% | Only Grok 4.20 measured |
| MedQA (Vals AI) | Not measured | 94.5% | Only Grok 4.20 measured |
| MedCode (Vals AI) | 51.5% | 32.2% | Claude Haiku 5.5 |
| SimpleQA Verified | 23.8% | Not measured | Only Claude Haiku 5.5 measured |
| Neither model has a result on the other 9 boards here | |||
Web designLike-for-likePublic boards, shown as published, kept out of the index60
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| Design Arena: websites | 1262 | 1229 | Claude Haiku 5.5 |
| Design Arena: UI components | 1291 | 1200 | Claude Haiku 5.5 |
| Design Arena: game dev | 1313 | 1193 | Claude Haiku 5.5 |
| Design Arena: data viz | 1294 | 1217 | Claude Haiku 5.5 |
| Design Arena: 3D | 1342 | 1193 | Claude Haiku 5.5 |
| Design Arena: all code categories | 1278 | 1221 | Claude Haiku 5.5 |
| Design Arena: SVG | Not measured | 1165 | Only Grok 4.20 measured |
| Design Arena: ASCII art | Not measured | 1175 | Only Grok 4.20 measured |
| LiveBench design | 83.1% | Not measured | Only Claude Haiku 5.5 measured |
GeneralLike-for-likePublic boards, shown as published, kept out of the index50
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | Leader |
|---|---|---|---|
| Vals Index | 58.7% | Not measured | Only Claude Haiku 5.5 measured |
| LiveBench data analysis | 53.1% | Not measured | Only Claude Haiku 5.5 measured |
| Public Benefits Bench (Vals AI) | 63.2% | Not measured | Only Claude Haiku 5.5 measured |
| SAGE (Vals AI) | 45.8% | 38.2% | Claude Haiku 5.5 |
| MedScribe (Vals AI) | 89.5% | 63.4% | Claude Haiku 5.5 |
| Vals Multimodal Index | Not measured | 39.1% | Only Grok 4.20 measured |
| IFBench | Not measured | 81.2% | Only Grok 4.20 measured |
| Long context reasoning (AA-LCR) | 82.7% | 69.0% | Claude Haiku 5.5 |
| DTBench | 95.0% | 90.1% | Claude Haiku 5.5 |
| LMCA | 51.1% | 38.7% | Claude Haiku 5.5 |
| CL-bench | Not measured | 22.2% | Only Grok 4.20 measured |
| Furniture assembly | 47.5% | Not measured | Only Claude Haiku 5.5 measured |
| CL-bench Life | Not measured | 11.9% | Only Grok 4.20 measured |
| Neither model has a result on the other 33 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Claude Haiku 5.5 | Grok 4.20 | 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 | SpaceXAI Grok 4.20 |
|---|---|---|
| Maker | Anthropic | xAI |
| Weights | No public weights listed | No public weights listed |
| Released | October 7, 2026 | March 31, 2026 |
| Context window | 1M | 2M |
| Input $/Mtok | $0.10 | $1.25 |
| Output $/Mtok | $0.50 | $2.50 |
| Coding index | 70.7 | 33.3 |
| Independent measurements behind it | 5 | 2 |
| Cost per solved task | Not computable | $0.72 |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Claude Haiku 5.5 better than Grok 4.20 for coding?
Pick Claude Haiku 5.5 for coding: it scores 70.7 on the Vibe Coding Index against 33.3 for Grok 4.20. Pick Grok 4.20 for a larger context window (2M against 1M tokens) and work measured by Vending-Bench 2 or ARC-AGI, where it leads.
Which is cheaper, Claude Haiku 5.5 or Grok 4.20?
Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens; Grok 4.20 costs $1.25 and $2.50. Cost per solved task is known for Grok 4.20 only.
Which has the bigger context window, Claude Haiku 5.5 or Grok 4.20?
Grok 4.20, with 2M tokens against 1M.
How many benchmarks is the Claude Haiku 5.5 vs Grok 4.20 comparison based on?
24 benchmarks have a result for both models as of October 22, 2026. Claude Haiku 5.5's coding index rests on 5 independent benchmarks and Grok 4.20's on 2. Where only one model has a result, the page says so instead of naming a winner.