AnthropicClaude Haiku 5.5
Claude Haiku 5.5 is an Anthropic proprietary reasoning model, released 7 Oct 2026. Its best configuration (xhigh reasoning effort) ranks #96 of 775 on the BenchLeader Index at 61.5 ±4.9, in the upper half. The ± is the point: 197 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in composite (76) and lowest in coding (57). At $0.200 per million tokens blended it is cheaper than most ranked models. Output speed of 192 tokens per second puts it in the fastest quarter, with a first answer in 59.5 s. It has been measured at 7 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 7 Oct 2026.
- Blended price
- $0.200/M
- $0.100 in · $0.500 out
- Output speed
- 192 tok/s
- measured by Artificial Analysis
- First answer
- 60 s
- first token 1.05 s
- Context
- 1M
- Full answer
- 62 s
- median, reasoning included
- Cost per run
- $0.124
- one full Intelligence Index run
- Released
- 7 Oct 2026
- Overall index62
- Reasoning68
- Coding57
- Maths67
- Knowledge64
- Long context64
- Composite76
Versions
Anthropic has shipped 6 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
| Model | Released | Index | Rank |
|---|---|---|---|
| Claude Haiku 5.5xhighthis page | 7 Oct 2026 | 61.5 | #96 |
| Claude Haiku 4.5no reasoning | 15 Oct 2025 | 48.5 | #419 |
| Claude 3.5 Haiku | 4 Nov 2024 | 39.7 | #666 |
| Claude 3 Haiku | 7 Mar 2024 | 37.0 | #727 |
| Claude Haiku 3 | – | – | – |
| Claude Haiku 4thinking | – | – | – |
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | – | – | 155 tok/s | 1.05 s | $0.0002 | |
| low | 54.8 | #239 | 171 tok/s | 8.39 s | $0.0002 | Coding 53 · Composite 63 · Knowledge 63 · Long context 61 · Reasoning 45 |
| medium | 56.6 | #196 | 178 tok/s | 11 s | $0.0002 | Coding 53 · Composite 68 · Knowledge 64 · Long context 64 · Reasoning 49 |
| high | 59.4 | #129 | 179 tok/s | 23 s | $0.0002 | Agents & tools 52 · Coding 59 · Composite 72 · Knowledge 64 · Long context 64 · Maths 66 · Reasoning 59 |
| xhighbest | 61.5 | #96 | 192 tok/s | 60 s | $0.0002 | Coding 57 · Composite 76 · Knowledge 64 · Long context 64 · Maths 67 · Reasoning 68 |
| max | 60.6 | #109 | 234 tok/s | 245 s | $0.0002 | Agents & tools 56 · Coding 61 · Composite 78 · Knowledge 51 · Long context 67 · Maths 65 · Reasoning 63 |
| not stated | – | – | 155 tok/s | 1.05 s | $0.0002 | Coding 43 · Composite 59 · Instruction following 64 · Reasoning 60 |
What more thinking costs
Turning the effort up buys index points and multiplies the bill. Here max costs 8.7× low for +5.8 on the index — 1.9 points per doubling of spend. Compare that with other models.
How its index has moved
8 Oct 2026 to 20 Oct 2026- xhigh
- max
- high
- medium
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | low | medium | high | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|---|---|---|---|
| GPQA Diamond | – | – | – | – | – | 89.7%#51 | – | Epoch AI Benchmarking Hub | |
| Humanity's Last Exameffort not stated | – | – | – | – | – | – | 30.1%#13 | Scale AI / CAIS | |
| LiveBench 2.0 Reasoningnot in indexeffort not stated | – | – | – | – | – | – | 82.4%#36 | LiveBench | |
| LiveBench 1.0 Reasoningnot in index | – | – | – | – | 81.2%#50 | 84.6%#40 | – | LiveBench | |
| Humanity's Last Exam (AA)not in index | – | 27.0%#174 | 33.8%#128 | 37.3%#104 | 42.7%#63 | 44.4%#51 | – | Artificial Analysis | |
| ARC-AGI-1 | – | 50.3%#167 | 57.3%#156 | 81.5%#110 | 87.5%#85 | 89.0%#79 | – | ARC Prize | |
| ARC-AGI-2 | – | 2.8%#184 | 9.4%#151 | 30.7%#123 | 61.3%#77 | – | – | ARC Prize | |
| ARC-AGI-3 | – | – | – | – | – | 2.8%#30 | – | ARC Prize | |
| CritPt | – | 9.1%#126 | 12.9%#104 | 18.6%#67 | 22.6%#54 | 18.9%#66 | – | Artificial Analysis | |
| MysteryMechanismnot in indexeffort not stated | – | – | – | – | – | – | 35.1%#10 | Vals AI | |
| Chess Puzzlesnot in index | – | 24.0%#79 | – | – | – | 55.0%#16 | – | Epoch AI Benchmarking Hub | |
| Mystery Game Puzzlesnot in index | – | – | – | – | – | 30.0%#33 | – | Epoch AI Benchmarking Hub | |
| LMCAnot in index | 39.3%#109 | 44.8%#92 | 45.6%#87 | 45.2%#89 | 48.0%#78 | 51.1%#62 | – | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | 75.2%#149 | 81.1%#126 | 84.3%#109 | 86.7%#104 | 87.2%#101 | 95.0%#45 | – | Epoch AI Benchmarking Hub |
Coding
| Benchmark | low | medium | high | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| SciCode | 49.2%#116 | 49.0%#118 | 48.7%#120 | 51.7%#91 | 55.0%#52 | – | SciCode | |
| FrontierCode | – | – | – | – | 46.4%#15 | – | Cognition | |
| Arena WebDev | – | – | 1586#31 | – | – | – | Arena (LMArena) | |
| LiveBench 2.0 Codingnot in indexeffort not stated | – | – | – | – | – | 91.4%#10 | LiveBench | |
| LiveBench 1.0 Codingnot in index | – | – | – | 76.4%#46 | 77.9%#35 | – | LiveBench | |
| SciCode (AA)not in index | 49.2%#116 | 49.0%#119 | 48.7%#121 | 51.7%#90 | 55.0%#54 | – | Artificial Analysis | |
| IOIeffort not stated | – | – | – | – | – | 47.3%#28 | Vals AI | |
| Code Migrationnot in indexeffort not stated | – | – | – | – | – | 44.6%#19 | Vals AI | |
| ProgramBenchnot in indexeffort not stated | – | – | – | – | – | 2.5%#7 | Vals AI | |
| Vibe Code Bench 1-100not in indexeffort not stated | – | – | – | – | – | 20.4%#10 | Vals AI | |
| Vibe Code Bench v1.1not in indexeffort not stated | – | – | – | – | – | 90.4%#3 | Vals AI | |
| CursorBenchnot in index | 30.9%#64 | 36.9%#46 | 42.3%#31 | 44.3%#26 | 48.4%#15 | – | Cursor | |
| FrontierSWEnot in index | – | – | – | – | 43.8%#8 | – | Epoch AI Benchmarking Hub | |
| Terminal-Bench 4.0 (AA)not in index | 12.6%#77 | 15.2%#69 | 21.7%#58 | 29.3%#47 | 32.8%#41 | – | Artificial Analysis | |
| FrontierCode 1.1 (AA)not in index | 29.3%#37 | 40.4%#25 | 43.6%#15 | 45.5%#10 | 41.8%#22 | – | Artificial Analysis |
Agents & tools
| Benchmark | low | medium | high | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| APEX-Agents | – | – | – | – | 56.1%#18 | – | Mercor | |
| Arena Agent | – | – | 1.5#24 | – | – | – | Arena (LMArena) | |
| LiveBench Long Agenticnot in indexeffort not stated | – | – | – | – | – | 57.6%#9 | LiveBench | |
| LiveBench Browsernot in indexeffort not stated | – | – | – | – | – | 80.0%#16 | LiveBench | |
| LiveBench Agentic Codingnot in index | – | – | – | 51.4%#40 | 57.2%#22 | – | LiveBench | |
| GDPval-AA v2.1not in index | 31.1%#134 | 38.7%#105 | 45.9%#71 | 50.6%#48 | 55.9%#25 | – | Artificial Analysis | |
| Analyst Agent (AA)not in index | – | – | – | – | 53.8%#6 | – | Artificial Analysis | |
| CyberBenchnot in indexeffort not stated | – | – | – | – | – | 75.5%#6 | Vals AI | |
| Finance Agent v2not in indexeffort not stated | – | – | – | – | – | 54.1%#27 | Vals AI | |
| Harvey's Legal Agent Benchmarknot in indexeffort not stated | – | – | – | – | – | 4.2%#20 | Vals AI | |
| Legal Research Benchnot in indexeffort not stated | – | – | – | – | – | 43.3%#22 | Vals AI | |
| SREBenchnot in indexeffort not stated | – | – | – | – | – | 5.7%#16 | Vals AI | |
| Tax Agent Benchnot in indexeffort not stated | – | – | – | – | – | 31.9%#21 | Vals AI | |
| Terminal-Bench 4.0 (Vals)not in indexeffort not stated | – | – | – | – | – | 36.4%#15 | Vals AI | |
| Terminal-Bench Sciencenot in indexeffort not stated | – | – | – | – | – | 10.0%#15 | Vals AI | |
| Vending-Bench 2not in indexeffort not stated | – | – | – | – | – | 3602.5#46 | Andon Labs | |
| AutomationBenchnot in index | 22.9%#128 | 28.6%#116 | 33.7%#111 | 36.0%#104 | 35.4%#107 | – | Artificial Analysis | |
| GDP.pdf (AA)not in index | 11.2%#117 | 15.2%#96 | 17.2%#82 | 18.2%#75 | 20.8%#58 | – | Artificial Analysis | |
| Harvey LAB-AA v1.1not in index | 1.7%#16 | 0.6%#20 | 1.1%#18 | 2.8%#14 | 0.6%#20 | – | Harvey | |
| AA-Briefcase v1.1not in index | 1112#48 | – | – | 1532#26 | 1577#17 | – | Artificial Analysis | |
| Terminal-Bench-Science 0.1 (AA)not in index | 1.9%#45 | 7.1%#33 | 10.0%#26 | 18.1%#17 | 20.0%#16 | – | Artificial Analysis | |
| FACTS Searchnot in indexeffort not stated | – | – | – | – | – | 77.5%#16 |
Maths
| Benchmark | high | xhigh | max | Source | Trend |
|---|---|---|---|---|---|
| FrontierMath Tiers 1–3 | – | – | 75.1%#20 | Epoch AI Benchmarking Hub | |
| FrontierMath Tier 4 | – | – | 46.3%#27 | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 97.2%#36 | 98.9%#18 | – | Epoch AI Benchmarking Hub | |
| ProofBench | – | – | 89.0%#9 | Vals AI | |
| LiveBench Mathematicsnot in index | – | 92.8%#27 | 86.4%#50 | LiveBench |
Knowledge
| Benchmark | low | medium | high | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| SimpleQA Verified | – | – | – | – | 23.8%#70 | – | Epoch AI Benchmarking Hub | |
| LiveBench 2.0 Data Analysisnot in indexeffort not stated | – | – | – | – | – | 53.1%#47 | LiveBench | |
| LiveBench Researchnot in indexeffort not stated | – | – | – | – | – | 81.9%#15 | LiveBench | |
| LiveBench 1.0 Data Analysisnot in index | – | – | – | 72.8%#44 | 51.8%#66 | – | LiveBench | |
| AA-Omniscience | 3.3#115 | 4.5#110 | 5.8#103 | 6.0#101 | 10.7#93 | – | Artificial Analysis | |
| BioMysteryBenchnot in indexeffort not stated | – | – | – | – | – | 70.4%#12 | Vals AI | |
| Excel Modeling Benchmarknot in indexeffort not stated | – | – | – | – | – | 59.0%#37 | Vals AI | |
| MedCodenot in indexeffort not stated | – | – | – | – | – | 51.5%#15 | Vals AI | |
| MedScribenot in indexeffort not stated | – | – | – | – | – | 89.5%#6 | Vals AI | |
| Public Benefits Benchnot in indexeffort not stated | – | – | – | – | – | 63.2%#24 | Vals AI | |
| EBR-benchnot in index | – | – | – | – | 43.8%#9 | – | Epoch AI Benchmarking Hub | |
| AA-Omniscience: accuracynot in index | 33.3%#161 | 34.0%#157 | 34.8%#154 | 35.0%#150 | 36.4%#147 | – | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 55.1%#99 | 55.2%#97 | 55.5%#95 | 55.6%#94 | 59.6%#79 | – | Artificial Analysis |
Instruction following
| Benchmark | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|
| LiveBench Languagenot in index | 63.5%#62 | 62.6%#64 | – | LiveBench | |
| LiveBench Instruction Followingnot in index | 66.5%#40 | 54.7%#65 | – | LiveBench | |
| MultiChallengeeffort not stated | – | – | 67.7%#6 | Scale AI SEAL |
Multimodal
| Benchmark | max | not stated | Source | Trend |
|---|---|---|---|---|
| LiveBench Designnot in indexeffort not stated | – | 83.1%#6 | LiveBench | |
| SAGEnot in indexeffort not stated | – | 45.8%#38 | Vals AI | |
| Blueprint-Bench 2not in indexeffort not stated | – | 33.0%#13 | Epoch AI Benchmarking Hub | |
| AA-3Dnot in index | 997#10 | – | Artificial Analysis |
Long context
| Benchmark | low | medium | high | xhigh | max | Source | Trend |
|---|---|---|---|---|---|---|---|
| AA-LCR | 72.0%#188 | 77.3%#131 | 77.3%#131 | 78.3%#111 | 82.7%#35 | Artificial Analysis | |
| MLCRnot in index | – | – | – | – | 63.3%#5 | Artificial Analysis |
Composite
| Benchmark | low | medium | high | xhigh | max | not stated | Source | Trend |
|---|---|---|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | – | – | – | – | 159.3#16 | Epoch AI Benchmarking Hub | |
| LiveBench 2.0effort not stated | – | – | – | – | – | 75.6%#15 | LiveBench | |
| LiveBench 1.0not in index | – | – | – | 72.1%#50 | 67.9%#59 | – | LiveBench | |
| AA Intelligence Index v4.3.2 | 29.4#127 | 34.5#92 | 37.8#81 | 41.3#57 | 43.4#46 | – | Artificial Analysis | |
| Vals Indexnot in indexeffort not stated | – | – | – | – | – | 58.7#10 | Vals AI |
Safety & honesty
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| FORTRESSnot in index | 9.5%#66 | Scale AI SEAL |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Google Vertex | 191 tok/s | 1.80 s | $0.110 | $0.550 | 1M | – |
| Google Vertex | 181 tok/s | 1.74 s | $0.100 | $0.500 | 1M | – |
| Amazon Bedrock | 177 tok/s | 0.87 s | $0.110 | $0.550 | 1M | – |
| Amazon Bedrock | 170 tok/s | 1.44 s | $0.110 | $0.550 | 1M | – |
| Claude Platform on AWS | 158 tok/s | 1.26 s | $0.100 | $0.500 | 1M | – |
| Google Vertex | 158 tok/s | 0.47 s | $0.110 | $0.550 | 1M | – |
| Anthropic | 147 tok/s | 0.83 s | $0.100 | $0.500 | 1M | – |
| Azure | 138 tok/s | 1.29 s | $0.100 | $0.500 | 1M | – |
| Amazon Bedrock | 122 tok/s | 2.48 s | $0.100 | $0.500 | 1M | – |
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache at $0.010 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0002 | $0.0002 | 1.0 min |
| Summarize a 30-page report | 12,000 / 600 | $0.0015 | $0.0007 | 1.0 min |
| Code edit | 6,000 / 1,500 | $0.0014 | $0.0009 | 1.1 min |
| Agentic coding session | 60,000 / 4,000 | $0.0080 | $0.0040 | 1.3 min |
| Structured extraction | 2,000 / 200 | $0.0003 | $0.0002 | 1.0 min |
More from Anthropic
Compare Claude Haiku 5.5 with…
- Claude Haiku 5.5 vs Claude Opus 5.5
- Claude Haiku 5.5 vs Gemini 4 Argon
- Claude Haiku 5.5 vs Muse Wonder Full
- Claude Haiku 5.5 vs GPT-6 Astra
- Claude Haiku 5.5 vs Claude Fable 5.1
- Claude Haiku 5.5 vs GPT-6.1 Sol
- Claude Haiku 5.5 vs Claude Sonnet 5.5
- Claude Haiku 5.5 vs Claude Fable 5
- Claude Haiku 5.5 vs Claude Opus 5
- Claude Haiku 5.5 vs GPT-5.6 Sol
See also
Data as of 21 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Claude Haiku 5.5: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/claude-haiku-5-5, data as of 21 Oct 2026.