Claude Haiku 4.5 (latest)

Claude Haiku 4.5 is Anthropic’s fastest and most efficient model, delivering near-frontier intelligence at a fraction of the cost and latency of larger Claude models. Matching Claude Sonnet 4’s performance across reasoning, coding, and computer-use tasks, Haiku 4.5 brings frontier-level capability to real-time and high-volume applications.

It introduces extended thinking to the Haiku line; enabling controllable reasoning depth, summarized or interleaved thought output, and tool-assisted workflows with full support for coding, bash, web search, and computer-use tools. Scoring >73% on SWE-bench Verified, Haiku 4.5 ranks among the world’s best coding models while maintaining exceptional responsiveness for sub-agents, parallelized execution, and scaled deployment.

anthropic/claude-haiku-4-5

Organization: Anthropic
Family: claude-haiku
Providers: 37
Context: 200,000
Output limit: 64,000
Knowledge: 2025-02-28
Release: 2025-10-15
Updated: —
Weights: Closed
Input: text, image/jpeg, image/png, image/gif, image/webp, application/pdf
Output: text
Capabilities: —

Providers

Compare every cataloged offer for this model, including its provider, route, token limits, price, and capabilities.

48 offers · USD per million tokens

Provider	Model ID	Context	Output	Input / output · 1M	Reasoning	Tools	Structured	Details
302.AI	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
AIHubMix	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
AI/ML API	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
Anthropic	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
AnyAPI	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Cloudflare AI Gateway	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
Cortecs	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
CrossModel	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Databricks	databricks-claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
DevPass (LLM Gateway)	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
DigitalOcean	anthropic-claude-haiku-4.5	200,000	8,192	— / —	—	—	—	View
FrogBot	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
GitHub Copilot	claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
Helicone	claude-4.5-haiku	200,000	8,192	— / —	—	—	—	View
Impossibl	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Kilo Gateway	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
LLM Gateway	vertex-anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
LLM Gateway	aws-bedrock/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
LLM Gateway	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Model Oracle AI	claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
NanoGPT	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
NanoGPT	anthropic/claude-haiku-4.5:thinking	200,000	64,000	— / —	—	—	—	View
NEAR AI Cloud	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Neon	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Ofox	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
OpenCode	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Amazon Bedrockvia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.10 / $5.50	Yes	Yes	Yes	View
Amazon Bedrockvia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.00 / $5.00	Yes	Yes	Yes	View
Amazon Bedrockvia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.10 / $5.50	Yes	Yes	Yes	View
Anthropicvia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.00 / $5.00	Yes	Yes	Yes	View
Azurevia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.00 / $5.00	Yes	Yes	Yes	View
Googlevia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.10 / $5.50	Yes	Yes	Yes	View
Googlevia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.10 / $5.50	Yes	Yes	Yes	View
Googlevia OpenRouter	anthropic/claude-haiku-4.5 unknown	200,000	64,000	$1.00 / $5.00	Yes	Yes	Yes	View
Opper	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
OrcaRouter	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
Perplexity Agent	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Pioneer	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Poe	anthropic/claude-haiku-4.5	192,000	64,000	— / —	—	—	—	View
Qiniu	claude-4.5-haiku	200,000	64,000	— / —	—	—	—	View
Requesty	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Requesty	claude-haiku-4-5@eu	200,000	64,000	— / —	—	—	—	View
SAP AI Core	anthropic--claude-4.5-haiku	200,000	64,000	— / —	—	—	—	View
Snowflake Cortex	claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Tempr Gateway	anthropic/claude-haiku-4-5	200,000	64,000	— / —	—	—	—	View
Vercel AI Gateway	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View
Xpersona	claude-haiku-4-5	200,000	128,000	— / —	—	—	—	View
ZenMux	anthropic/claude-haiku-4.5	200,000	64,000	— / —	—	—	—	View

Prices and limits apply to each offer. A dash means the value is not provided. OpenRouter offers show the hosting provider and routing channel separately.

Pricing

Compare token rates and cache charges across available routes and hosting configurations.

USD per million tokens

Input from$1.00Per 1M input tokens

Output from$5.00Per 1M output tokens

Available offers48Across 37 providers

Lowest listed input and output prices across the offers above; they may belong to different providers. These are base rates, before conditional pricing, discounts or additional fees.

Endpoint price comparisonLatest observed snapshot · 8 lowest-priced configurations

InputOutput

Amazon Bedrockamazon-bedrock/global

$1.00$5.00

Anthropicanthropic

$1.00$5.00

Azureazure/global

$1.00$5.00

Googlegoogle-vertex/global

$1.00$5.00

Amazon Bedrockamazon-bedrock/us

$1.10$5.50

Amazon Bedrockamazon-bedrock/eu-west-1

$1.10$5.50

Googlegoogle-vertex/us-east5

$1.10$5.50

Googlegoogle-vertex/europe

$1.10$5.50

anthropic/claude-haiku-4.5Observed Oct 10, 2026, 11:31 AM UTC

Input: $1.00
Output: $5.00
Cache read: $0.10
Cache write: $1.25
Cache write · 1 hour: $2.00

Additional fees and pricing conditions

web search: $0.01 / search

Hosting prices through OpenRouter All endpoints and cache rates

Endpoint	Input / 1M	Output / 1M	Cache read / 1M	Cache write / 1M
Amazon Bedrockamazon-bedrock/us	$1.10	$5.50	$0.11	$1.375
Amazon Bedrockamazon-bedrock/global	$1.00	$5.00	$0.10	$1.25
Amazon Bedrockamazon-bedrock/eu-west-1	$1.10	$5.50	$0.11	$1.375
Anthropicanthropic	$1.00	$5.00	$0.10	$1.25
Azureazure/global	$1.00	$5.00	$0.10	$1.25
Googlegoogle-vertex/us-east5	$1.10	$5.50	$0.11	$1.375
Googlegoogle-vertex/europe	$1.10	$5.50	$0.11	$1.375
Googlegoogle-vertex/global	$1.00	$5.00	$0.10	$1.25

Each row is a hosting configuration. Open an offer in Providers for its full pricing conditions and additional fees. Endpoint observations may differ from the routing catalog's base price.

Performance

Review reported latency and output-throughput percentiles from the latest observed rolling window.

Latest 30-minute observation

Time to first token · milliseconds

Latency distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Amazon Bedrockamazon-bedrock/us

654–1,043 ms

Amazon Bedrockamazon-bedrock/global

833.5–4,997.71 ms

Amazon Bedrockamazon-bedrock/eu-west-1

508–2,112.3 ms

Anthropicanthropic

551–1,617.45 ms

Azureazure/global

354–1,780.8 ms

Googlegoogle-vertex/us-east5

2,523–2,588.76 ms

All latency percentiles 8 endpoints

Endpoint	P50	P75	P90	P99
Amazon Bedrockamazon-bedrock/us	654 ms	728 ms	802.9 ms	1,043 ms
Amazon Bedrockamazon-bedrock/global	833.5 ms	1,178 ms	1,817.9 ms	4,997.71 ms
Amazon Bedrockamazon-bedrock/eu-west-1	508 ms	540 ms	643 ms	2,112.3 ms
Anthropicanthropic	551 ms	1,004 ms	1,156.9 ms	1,617.45 ms
Azureazure/global	354 ms	430.5 ms	1,305 ms	1,780.8 ms
Googlegoogle-vertex/us-east5	2,523 ms	2,551.25 ms	2,569.1 ms	2,588.76 ms
Googlegoogle-vertex/europe	744 ms	806.5 ms	949 ms	1,556.4 ms
Googlegoogle-vertex/global	544 ms	781.25 ms	2,265.1 ms	4,575.48 ms

Output throughput · tokens per second

Throughput distributionRolling 30-minute snapshot · 6 endpoints

P50P75P90P99

Amazon Bedrockamazon-bedrock/us

103–179.63 t/s

Amazon Bedrockamazon-bedrock/global

69–178.09 t/s

Amazon Bedrockamazon-bedrock/eu-west-1

85–162.08 t/s

Anthropicanthropic

73–140 t/s

Azureazure/global

93–130 t/s

Googlegoogle-vertex/us-east5

198–201.46 t/s

All throughput percentiles 8 endpoints

Endpoint	P50	P75	P90	P99
Amazon Bedrockamazon-bedrock/us	103 t/s	119 t/s	147.1 t/s	179.63 t/s
Amazon Bedrockamazon-bedrock/global	69 t/s	91 t/s	114 t/s	178.09 t/s
Amazon Bedrockamazon-bedrock/eu-west-1	85 t/s	92 t/s	104.8 t/s	162.08 t/s
Anthropicanthropic	73 t/s	91 t/s	104 t/s	140 t/s
Azureazure/global	93 t/s	101 t/s	110 t/s	130 t/s
Googlegoogle-vertex/us-east5	198 t/s	199 t/s	199 t/s	201.46 t/s
Googlegoogle-vertex/europe	80.5 t/s	85.75 t/s	126.2 t/s	136.43 t/s
Googlegoogle-vertex/global	81 t/s	91 t/s	101 t/s	119.39 t/s

Percentiles describe the observed request distribution. These are snapshots of a rolling window, not a historical time series.

Uptime

Compare successful request rates across the latest 5-minute, 30-minute, and 24-hour observations.

Successful requests by window

Availability by endpointCurrent overlapping windows · 8 endpoints

Endpoint5 minutes30 minutes24 hours

Amazon Bedrockamazon-bedrock/usObserved Oct 10, 2026, 11:31 AM UTC

100%

Amazon Bedrockamazon-bedrock/globalObserved Oct 10, 2026, 11:31 AM UTC

100%

Amazon Bedrockamazon-bedrock/eu-west-1Observed Oct 10, 2026, 11:31 AM UTC

—

100%

AnthropicanthropicObserved Oct 10, 2026, 11:31 AM UTC

100%

Azureazure/globalObserved Oct 10, 2026, 11:31 AM UTC

100%

Googlegoogle-vertex/us-east5Observed Oct 10, 2026, 11:31 AM UTC

—

100%

Googlegoogle-vertex/europeObserved Oct 10, 2026, 11:31 AM UTC

100%

Googlegoogle-vertex/globalObserved Oct 10, 2026, 11:31 AM UTC

—

100%

99.68%

Reported by OpenRouter at the time of collection; rate-limited requests are excluded. A dash means no measurement was reported. This is not a live availability check.

Benchmarks

Review attributed evaluation results on each source's original scale.

External evaluations

anthropic/claude-haiku-4.5Observed Oct 10, 2026, 11:31 AM UTC

Benchmark profileArtificial Analysis · shared source scale from 0 to 50

Coding

43.9

Agentic

Intelligence

16.9

Design Arena	Category	Elo	Win rate	Rank
models	3d	1,077	41.1%	93
models	asciiart	1,145	49.3%	47
models	codecategories	1,119	44.9%	89
models	dataviz	1,128	45.6%	87
models	gamedev	1,104	44.7%	88
models	svg	1,025	39.1%	79
models	uicomponent	1,103	42.6%	88
models	website	1,127	45.1%	92

Sources: Artificial Analysis and Design Arena, via OpenRouter. Scores retain their original scales and are not OpenAI Suite ratings.

Detailed evaluations

Evaluation scoresAccuracy or item F1 · higher is better

gpqa diamondAnthropic: Claude Haiku 4.5

73.156%

tau bench verified airlineAnthropic: Claude Haiku 4.5

65.667%

Evaluation	Configuration	Source	Results	Details
models · codecategories	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,104 Elo	Details
models · website	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,114 Elo	Details
models · svg	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,015 Elo	Details
models · 3d	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,052 Elo	Details
models · asciiart	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,135 Elo	Details
gpqa diamond	Anthropic: Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	OpenRouter	73.156%	Details
tau bench verified airline	Anthropic: Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	OpenRouter	65.667%	Details
Composite indices	Claude 4.5 Haiku (Reasoning)anthropic/claude-4.5-haiku-20251001	Artificial Analysis	Intelligence 7.4 · Coding 43.9 · Agentic 8	Details
models · dataviz	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,122 Elo	Details
models · gamedev	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,087 Elo	Details
models · uicomponent	Claude Haiku 4.5anthropic/claude-4.5-haiku-20251001	Design Arena	1,090 Elo	Details

Sources: Artificial Analysis, Design Arena and OpenRouter. Evaluation configurations and scales differ; results are not interchangeable. Updated Nov 14, 2026.

Apps & session costs

Compare observed median session costs for applications using this model, grouped by conversation length.

30-day sample · through Nov 1, 2026

Application	Model route	Turns	Median cost / session	Details
Hermes Agent	anthropic/claude-4.5-haiku-20251001	10–49 turns	$0.178342	Details
Hermes Agent	anthropic/claude-4.5-haiku-20251001	50+ turns	$1.138054	Details
Claude Code	anthropic/claude-4.5-haiku-20251001	10–49 turns	$0.233418	Details
Kilo Code	anthropic/claude-4.5-haiku-20251001	1 turn	$0.003832	Details
Hermes Agent	anthropic/claude-4.5-haiku-20251001	2–9 turns	$0.034941	Details
Claude Code	anthropic/claude-4.5-haiku-20251001	2–9 turns	$0.04327	Details
Hermes Agent	anthropic/claude-4.5-haiku-20251001	1 turn	$0.013765	Details
Claude Code	anthropic/claude-4.5-haiku-20251001	50+ turns	$1.200103	Details
Claude Code	anthropic/claude-4.5-haiku-20251001	1 turn	$0.013451	Details

Source: OpenRouter, as of Nov 9, 2026. Published under CC BY 4.0. These are observed costs per session, not token prices or a forecast for your workload. Apps are compared separately.

Activity

Daily prompt and completion tokens reported for this model's routes in OpenRouter's top 50. Missing days have no published value.

Last 90 complete UTC days

Daily token usage772,931,181,294 reported tokens · 22 days with observations

47B0

2026-08-16
24,618,337,189 tokens

2026-08-17
33,434,462,228 tokens

2026-08-18
38,644,088,357 tokens

2026-08-19
39,000,579,353 tokens

2026-08-20
38,397,190,365 tokens

2026-08-21
33,961,091,713 tokens

2026-08-22
24,029,338,908 tokens

2026-08-23
25,484,277,824 tokens

2026-08-24
37,661,688,851 tokens

2026-08-25
38,713,153,906 tokens

2026-08-26
39,875,874,160 tokens

2026-08-27
38,107,037,578 tokens

2026-08-28
34,634,271,797 tokens

2026-08-29
26,888,785,597 tokens

2026-08-30
26,690,446,152 tokens

2026-08-31
35,778,927,890 tokens

2026-09-01
41,138,215,828 tokens

2026-09-02
43,345,538,431 tokens

2026-09-03
37,036,342,538 tokens

2026-09-04
33,185,651,876 tokens

2026-09-08
46,532,610,084 tokens

2026-09-09
35,773,270,669 tokens

Aug 16, 2026Nov 13, 2026

Daily observations · 22 days

Date (UTC)	Reported tokens	Reported routes
Sep 9, 2026	35,773,270,669	1
Sep 8, 2026	46,532,610,084	1
Sep 4, 2026	33,185,651,876	1
Sep 3, 2026	37,036,342,538	1
Sep 2, 2026	43,345,538,431	1
Sep 1, 2026	41,138,215,828	1
Aug 31, 2026	35,778,927,890	1
Aug 30, 2026	26,690,446,152	1
Aug 29, 2026	26,888,785,597	1
Aug 28, 2026	34,634,271,797	1
Aug 27, 2026	38,107,037,578	1
Aug 26, 2026	39,875,874,160	1
Aug 25, 2026	38,713,153,906	1
Aug 24, 2026	37,661,688,851	1
Aug 23, 2026	25,484,277,824	1
Aug 22, 2026	24,029,338,908	1
Aug 21, 2026	33,961,091,713	1
Aug 20, 2026	38,397,190,365	1
Aug 19, 2026	39,000,579,353	1
Aug 18, 2026	38,644,088,357	1
Aug 17, 2026	33,434,462,228	1
Aug 16, 2026	24,618,337,189	1

Source: OpenRouter (openrouter.ai/rankings), as of Nov 14, 2026. CC BY 4.0. Totals include only individually published routes. Tokenizers differ by provider; the aggregated “other” category is never assigned to a model.

Technical details & API

Explore specifications, supported parameters, reasoning controls, and API identifiers for each OpenRouter route.

Current route specifications

anthropic/claude-haiku-4.5Observed Oct 10, 2026, 11:31 AM UTC

Context: 200,000
Maximum output: 64,000
Tokenizer: Claude
Input: text, image, file
Output: text
Moderation: Yes
Added to OpenRouter: 2025-10-15

Base URLhttps://openrouter.ai/api/v1

Model IDanthropic/claude-haiku-4.5

Version IDanthropic/claude-4.5-haiku-20251001

Reasoning

Required: No
Enabled by default: Not reported
Default effort: —
Supported efforts: —

Supported parameters

include_reasoning
max_completion_tokens
max_tokens
reasoning
response_format
stop
structured_outputs
temperature
tool_choice
tools
top_k
top_p

Frequently asked questions

Answers based on the model metadata and serving offers currently in the catalog.

How much does Claude Haiku 4.5 (latest) cost?

Listed rates start at $1.00 per million input tokens and $5.00 per million output tokens. Prices vary by provider and configuration; see Pricing for details.

Which providers offer Claude Haiku 4.5 (latest)?

There are 48 cataloged offers across 37 providers. Compare model IDs, prices and limits in Providers.

What is the context window?

The cataloged model context is 200,000 tokens. Each serving endpoint may apply a different limit.

Does it support tools and structured output?

Tool calling: Not reported. Structured output: Not reported. Support can vary by endpoint.

Catalog history · 3 recorded revisions

Metadata observations since this model was first cataloged. These are not model release versions.

Oct 19, 2026 · 5a1796c33966
Sep 14, 2026 · e28966f5b31b
Sep 12, 2026 · e6602649c58c