Skip to content

Anthropic · Chat

Claude Haiku 5.5

Anthropic 🇺🇸ChatClaude familyOne of its sources says the weights aren't publicly available.The maker currently lists 29 Sep 2027 as the deprecation date. Plans can change.
From Netmind's listing:
Claude Haiku 5.5 is the fastest and most cost-efficient model in the Claude family, built for high-volume, latency-sensitive work such as classification, routing, extraction, and subagent tasks. It offers a 1M token context window, up to 128K output tokens, adaptive thinking with selectable reasoning effort, multimodal input, native tool use, and prompt caching.
llminfo IQ
134.6
Maker's price
$0.20 blended / 1M
Max context
1M
Listings
89

Evidence and sources

llminfo IQ 134.6, ranked among chat models. From 230 benchmarks. Plenty of data behind it.How llminfo IQ is computedCoding IQ 123.6, 95% interval 118–129. From 17 benchmark families across 7 sources. Plenty of data behind it. Uses the model's best published results. It has its own scale, separate from llminfo IQ.How Coding IQ is computedGood value (62.1). Cheaper per task than its benchmark results predict. From 43 benchmarks with measured task costs. Plenty of data behind it.How Value for money worksYap score: tokens used per coding task vs. models with the same resultsTypical (49 ±9). Generates about as many tokens as models with similar coding results. Measured at default effort. Little data behind it, so treat it as a rough estimate. By effort: default 49 · low 49 · medium 53 · high 57 · max 62.Median 40,947 output tokens per task at default effort.How the Yap score worksReleased 7 Oct 2026Maker HQ: US (according to Anthropic's own website)
All rankings: #13 on OpenRouter today #13 on OpenRouter this month #17 on OpenRouter this week #38 on NanoGPT recentlySources: Anthropic API data, Heroku Managed Inference data, LiteLLM data, Models.dev data, NetMind.AI, llminfo researchPrimary data from Anthropic API from Heroku Managed Inference: primary mode from LiteLLM: deprecation date from Models.dev: weights availability from NetMind.AI: description from llminfo research: release date
Tags from its sources:instructtoolsvisionreasoningdocumentproprietaryfastcost-efficient

Advantages and trade-offs

Compared with 10 chat models, picked mainly by popularity and IQ

Trade-offs

No open weights

The weights aren't public, so you can't run it yourself.

We only make a claim when there's enough data on models of the same type. Updated 22 Nov, 06:07 UTC.Prices come from the model's maker where they sell it, otherwise their official partner, otherwise the market median. This comparison uses 9 maker's prices.

Input & output

as stated by the model's providers
Text
Images
Files
Text

Capabilities

Supported operations and provider-stated features

VisionReasoningFunction callingStructured outputPrompt cachingPDF inputComputer useWeb searchBatch API

Listings

 Anthropicclaude-haiku-5-5Sources: Anthropic API data, LiteLLM data, Models.dev data1M$0.10cache read $0.01batch $0.05$0.50cache write 5m $0.125cache write 1h $0.20batch $0.25
 Nous Portalanthropic/claude-haiku-5.5:batch · batchSource: Nous Portal Inference Gateway data1M$0.05cache read $0.005$0.25cache write $0.0625
 OpenRouterInference by Anthropicanthropic/claude-haiku-5.5:batch · batchSource: OpenRouter data docs1M$0.05cache read $0.005$0.25cache write 5m $0.0625cache write 1h $0.10
 Vercelanthropic/claude-haiku-5.5-batch · batchSource: Vercel AI Gateway data1M$0.05cache read $0.005$0.25
 Iteracomputeanthropic/claude-haiku-5-5Source: Models.dev data1M$0.05cache read $0.005$0.25cache write $0.0625
 Outerlabsclaude-haiku-5-5Source: Models.dev data1M$0.06cache read $0.006$0.30cache write $0.075
 Amazon Web Servicesanthropic.claude-haiku-5-5Sources: AWS Bedrock Runtime data, LiteLLM data, Models.dev data1M$0.10cache read $0.01$0.50cache write 5m $0.125cache write 1h $0.20
 Microsoft Azureazure_ai/claude-haiku-5-5Sources: LiteLLM data, Models.dev data1M$0.10cache read $0.01$0.50cache write 5m $0.125cache write 1h $0.20
 Googlevertex_ai/claude-haiku-5-5Source: LiteLLM data1M$0.10cache read $0.01batch $0.05$0.50cache write 5m $0.125cache write 1h $0.20batch $0.25
 302aiclaude-haiku-5-5Source: Models.dev data1M$0.10cache read $0.01$0.50cache write $0.125

10 listings without a price are hidden.

Included in plans

Copilot EnterprisegithubSubscription$39/moextra use billed past the allowancePlan page
Copilot ProgithubSubscription$10/moextra use billed past the allowancePlan page
Copilot Pro+githubSubscription$39/moextra use billed past the allowancePlan page
Copilot BusinessgithubSubscription$19/moextra use billed past the allowancePlan page
Copilot MaxgithubSubscription$100/moextra use billed past the allowancePlan page

Benchmarks

Named effort levels run from lowest to highest. Select a result to see its source details.

Human preference

Arena ELO47 related benchmarks · 47 results
Source: LMArena data · Nov 2026
BenchmarkHigh
Arena ELO
Chinese
Creative Writing
English
Show 43 more benchmarksHide 43 benchmarks
BenchmarkHigh
Exclude Ties
Expert
French
German
Hard Prompts
Hard Prompts English
Industry Business and Management and Financial Operations
Industry Entertainment and Sports and Media
Industry Life and Physical and Social Science
Industry Mathematical
Industry Medicine and Healthcare
Industry Software and It Services
Industry Writing and Literature and Language
Japanese
Korean
Longer Query
Multi Turn
Non English
Polish
Russian
Spanish
Style-Controlled
Text Factuality
Text Factuality - Chinese
Text Factuality - Creative Writing
Text Factuality - English
Text Factuality - Exclude Ties
Text Factuality - Expert
Text Factuality - Hard Prompts
Text Factuality - Hard Prompts English
Text Factuality - Industry Business and Management and Financial Operations
Text Factuality - Industry Entertainment and Sports and Media
Text Factuality - Industry Life and Physical and Social Science
Text Factuality - Industry Mathematical
Text Factuality - Industry Medicine and Healthcare
Text Factuality - Industry Software and It Services
Text Factuality - Industry Writing and Literature and Language
Text Factuality - Longer Query
Text Factuality - Multi Turn
Text Factuality - Non English
Text Factuality - Russian
Design Arena EloSource: Design Arena Leaderboards data · Nov 2026

Agentic

Arena AgentSource: LMArena data · Nov 2026
GDPval-AA v22 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Artificial Analysis · medium
Artificial Analysis · max
OSWorld4 related benchmarks · 4 results
2.0 · Strict
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 82-task offline subset · 1080p · 500 steps · max · 5 runs
2.0 · Partial
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
82-task offline subset · 1080p · 500 steps · max · 5 runs
2.1 (partial)
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
2.1 (strict)
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
PhysicianBench11 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Reasoning effort · adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 100 EHR tasks · five attempts/task
Source: Clinical Benchmarks data docs · Reported by the model's developer. The source relaying it did not verify it. claim · Oct 2026
Setting not stated
Setting not stated
Setting not stated
Setting not stated
Setting not stated
Show 1 more resultHide 1 result
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
SWE-Bench Pro2 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Source: OpenRouter benchmark exports
7 Oct 2026 version · 90-day results
Across 5 providers lowest 72.9% median 73.3% highest 74%OpenRouter's automatic routing 72.8%
Show results from 5 providers and OpenRouter's automatic routingHide results from 5 providers and OpenRouter's automatic routing
  • Amazon Bedrock
  • Google
  • Anthropic
  • Azure
  • Claude Platform on AWS
  • OpenRouter's automatic routing
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Claude Code --bare · max · 10 trials/task · no internet · safeguards stop trial
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Claude Code --bare · max · 10 trials/task · no internet · safeguards stop trial
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Coding

Arena ELOSource: LMArena data · Nov 2026
BenchCADSource: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
FrontierCode 1.13 related benchmarks · 7 results
FrontierCode 1.1
Source: FrontierCode data
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Extended
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Claude Code · max · mean@5
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Main
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Claude Code · xhigh · mean@5
Claude Code · max · mean@5
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Proximus harness · max · 34 tasks · mean@5
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
ValsSource: Vals AI

Reasoning

Chartography2 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
no tools · adaptive thinking · max · 5 runs · Gemini 3.5 Flash judge
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Humanity's Last Exam3 related benchmarks · 4 results
Humanity's Last Exam
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
no tools · auto thinking · 980K task budget
(no tools, text-only)
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
(with tools)
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
web search/fetch + programmatic tools + code · auto thinking · 980K task budget
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
MILU2 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 11-language average · adaptive thinking · max · 5 trials
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
OfficeQA2 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · preselected extracted text + code · max · 5 runs
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
OfficeQA Pro2 results
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 133-question subset · preselected extracted text + code · max · 5 runs
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

General

Multilingual

Benchmark Nusantara17 related benchmarks · 17 results
Show 11 more resultsHide 11 results
EKI EstonianSource: Keelemudelite mõõdupuu data docs · Oct 2026

Science

GPQA Diamond7 results
Source: OpenRouter benchmark exports
7 Oct 2026 version · 90-day results
Across 5 providers lowest 86.7% median 86.9% highest 87.1%OpenRouter's automatic routing 86.6%
Show results from 5 providers and OpenRouter's automatic routingHide results from 5 providers and OpenRouter's automatic routing
  • Amazon Bedrock
  • Anthropic
  • Claude Platform on AWS
  • Google
  • Azure
  • OpenRouter's automatic routing
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Math

Vision

Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated

Instruction following

Factuality

Knowledge

Global MMLU (GMMLU)adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 42-language average · adaptive thinking · max · single trial · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.

Other

Artificial Analysis9 related benchmarks · 10 results
Briefcase
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · v1.1 · Artificial Analysis · medium
v1.1 · Artificial Analysis · max
Show 3 more resultsHide 3 results
BenchCAD · Vision2Code voxel IoUadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 1,000-file subset · no tools · max · 5 runs · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Chartography · With toolsadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · container + crop tool · adaptive thinking · max · 5 runs · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
HealthBench3 related benchmarks · 11 results
HealthBench
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Reasoning effort · adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · length-adjusted · single run
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · length-adjusted · max effort · 5-run mean
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Professional
Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
adaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · raw · no tools · Opus 4.8 grader · max · 5 trials
length-adjusted · no tools · Opus 4.8 grader · max · 5 trials
Source: Clinical Benchmarks data docs · Reported by the model's developer. The source relaying it did not verify it. claim · Oct 2026
Setting not stated
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
Setting not stated
Professional (raw)
Source: LLM Stats data · Reported by the model's developer. The source relaying it did not verify it. claim
ProgramBench · Average Pass Rateadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · 166 golden tasks · mini-swe-agent · no 6-hour timeout · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
Protocol Troubleshootingadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
SingleCellBenchadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.
SpatialBench Verifiedadaptive thinking · max effort · default sampling · 5-trial mean · context ≤1M unless noted · Reported by Anthropic · via BenchBoard docs · Reported by the model's developer. The source relaying it did not verify it.

Data credits: Data by LLM Stats · Epoch AI Benchmarking Hub, CC BY 4.0 (licence) · LMArena, CC BY 4.0 (licence) · BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links. · Baromètre de l’écriture IA, Johannes Eva; public numeric ranking data, © 2026 Johannes Eva, no results licence stated · Benchmark Nusantara, ibahasa (M. Khoirul Huda), doi:10.5281/zenodo.21970463. Test results CC BY 4.0, attribution to ibahasa; aggregate results CSV only. (licence) · BullshitBench by Peter Gostev, MIT (licence) · Deep20Bench edition 1.1, Patrick Heusser. Results CC BY 4.0; native question scores and fraction-to-percent success rates. (licence) · Steel.dev Agent Leaderboards, MIT (licence) · Keelemudelite mõõdupuu: Independent LLM leaderboard for Estonian. Eesti Keele Instituut. https://moodupuu.eki.ee, CC BY 4.0. Latest per-benchmark results selected with partial-run flags retained. (licence) · OpenRouter public benchmark leaderboard and search-lane exports, CC BY 4.0. Fractional source scores converted to percentages; immutable dated snapshots. (licence) · Clinical Benchmarks (clinicalbenchmarks.ai), CC BY 4.0; fractions converted to percentages; source documents retain their own terms (licence) · LLM Chess by Maxim Saplin, Apache-2.0 (Copyright [2024] [Maxim Saplin]) (licence) · MineBench, MIT (licence) · WeirdML v3 by Håvard Tveit Ihle (Norwegian Defence Research Establishment). API costs supported by Epoch AI, METR and NDRE.

Also known as: anthropic-claude-5.5-haiku, anthropic.claude-haiku-5-5, anthropic/claude-haiku-5-5, anthropic/claude-haiku-5.5, anthropic/claude-haiku-5.5-batch, anthropic/claude-haiku-5.5:batch, apac.anthropic.claude-haiku-5-5, au.anthropic.claude-haiku-5-5, aws-bedrock/claude-haiku-5-5, azure_ai/claude-haiku-5-5, bedrock/us-gov-east-1/anthropic.claude-haiku-5-5, bedrock/us-gov-west-1/anthropic.claude-haiku-5-5, bedrock_mantle/anthropic.claude-haiku-5-5, bedrock_mantle/us-gov-west-1/anthropic.claude-haiku-5-5, ca.anthropic.claude-haiku-5-5, claude haiku 5.5, claude-haiku-5-5, claude-haiku-5-5@default, claude-haiku-5-5@eu, claude-haiku-5.5-high, databricks/databricks-claude-haiku-5-5, eu.anthropic.claude-haiku-5-5, github_copilot/claude-haiku-5.5, global.anthropic.claude-haiku-5-5, jp.anthropic.claude-haiku-5-5, opencode_go/claude-haiku-5-5, perplexity/anthropic/claude-haiku-5-5, snowflake/claude-haiku-5-5, us-gov.anthropic.claude-haiku-5-5, us.anthropic.claude-haiku-5-5, vertex-anthropic/claude-haiku-5-5, vertex_ai/claude-haiku-5-5, vertex_ai/claude-haiku-5-5@default

Like this one. Tap or hover any dotted value, label or icon to see where it comes from and how it is measured. open details when tapped or hovered. How this data works