How much does Claude Haiku 5.5 cost?

Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens, using the list price from Anthropic as of 26 Oct 2026. That is about $0.04 for 100 chats of about 1,000 tokens in and 500 tokens back. Confirm on the maker's site before you buy, because prices change.

What is the context window of Claude Haiku 5.5?

Claude Haiku 5.5 has a context window of 1M tokens, roughly 750,000 words of text read at once. A single answer can be up to 128K tokens long.

Is Claude Haiku 5.5 open source?

We list Claude Haiku 5.5 as closed: it is used through its maker's apps or API and its weights are not published (Epoch AI lists its access as "API access").

How good is Claude Haiku 5.5 compared with other AI models?

On the Epoch Capabilities Index, Claude Haiku 5.5 scores 159.3 (likely range 156.2 to 162.5), which ranks 16 of 270 models we track. On Arena's human-vote text board it is ranked 87 of 166. Scores close together are a tie in practice.

How does Claude Haiku 5.5 compare with GPT-5.4 Pro?

GPT-5.4 Pro scores higher on 5 of 6 shared quality measures, Claude Haiku 5.5 on 0. 1 measure is too close to call. Claude Haiku 5.5 costs about 100% less than GPT-5.4 Pro on a blended price of $0.20 against $67.50 per million tokens. See the full comparison for every measure.

How current are these numbers for Claude Haiku 5.5?

The newest data on this page is dated 26 Oct 2026. Scores come from Epoch AI and Arena, prices from the open LiteLLM price list.

AI Rankings

Claude Haiku 5.5: Score, Price and Context Window

Where Claude Haiku 5.5 from Anthropic ranks, what it costs and how much text it can read, in plain numbers with the sources.

Data updated 26 Oct 2026

Anthropic
Closed weights
Released Oct 2026

Its score range overlaps with GPT-5.4 Pro, the model just above it, so a small gap in rank is not a meaningful difference.

Best for

Demanding tasks: its overall Epoch score is in the top 10% of models we track.
Very long documents: its context window is 1M tokens, roughly 750,000 words.
Keeping costs low: it is among the cheapest quarter of models in the top half by score.

These lines are worked out from the numbers on this page, not from hands-on testing.

Key facts

Overall score: 159.3; Rank 16 of 270. Likely range 156.2 to 162.5.
Human preference: chat: 1425; Rank 87 of 166. 6,080 votes, at the "High" setting.
Human preference: coding: 1485; Rank 75 of 166. 1,370 votes, at the "High" setting.
Human preference: web pages: 1586; Rank 25 of 85. 3,917 votes, at the "High" setting.
Price per million tokens: $0.10 in / $0.50 out; Blended $0.20. Cached input $0.01. List price from Anthropic, checked 26 Oct 2026.
Context window: 1M tokens; About 750,000 words at once. One answer up to 128K tokens.
Weights: Closed; Used through the maker's apps or API (Epoch AI: API access).

Features listed by the price list: Image input, PDF input, Tool use, Reasoning, Web search, Prompt caching, Structured output.

Try itClaude in our AI tools directoryAnthropic's AI assistant for research, writing, code, files, and connected tasks, with free and paid plans.

How Claude Haiku 5.5 stacks up

Each strip shows every model we track as faint bars. The dot is this model. Better is always to the right, even for price.

Claude Haiku 5.5 compared with all models we trackOverall capability: 159.3. Better than 94% of the other 269 models. Human preference: chat: 1425. Better than 47% of the other 165 models. Human preference: coding: 1485. Better than 55% of the other 165 models. Human preference: web pages: 1586. Better than 71% of the other 84 models. Price: $0.20 per 1M. Cheaper than 80% of the other 200 priced models. Context window: 1M tokens. Larger context window than 72% of the other 198 models.

Overall capability159.3
Better than 94% of the other 269 modelsTop 10%
Epoch Capabilities Index: one score that combines many tests.
Human preference: chat1425
Better than 47% of the other 165 models
How often people pick its answers over another model's on Arena.
Human preference: coding1485
Better than 55% of the other 165 modelsUpper half
How often people pick its answers to coding questions on Arena.
Human preference: web pages1586
Better than 71% of the other 84 modelsUpper half
How often people pick the web pages and small apps it builds on Arena.
Price$0.20 per 1M
Cheaper than 80% of the other 200 priced modelsTop 25%
Blended price per million tokens. Lower is better, so the cheapest models sit on the right.
Context window1M tokens
Larger context window than 72% of the other 198 modelsUpper half
How much text it can read at once. Bigger suits long documents.

Compared with the other published models that have each measure. Data updated 26 Oct 2026.

Where Claude Haiku 5.5 ranks

Leaderboards it appears on, best placing first.

See all leaderboards

Test results

Each test checks one kind of skill. Place shows where it stands among the models that took that test.

OTIS Mock AIME
Maths competition problems in the style of the US AIME exam, from the 2024 and 2025 sets. Every answer is a whole number.
98.9%
Place 18 of 174
tied with 3 others
FrontierMath (Tiers 1-3)
Original, very hard maths problems written by mathematicians. Each has an exact answer and the problems are kept private so models cannot have memorised them. This is the Tiers 1 to 3 set, version 2.
75.1%
Place 20 of 87
GPQA Diamond
Very hard multiple-choice science questions in biology, physics and chemistry, written by experts and designed so a quick web search does not give the answer.
89.7%
Place 45 of 183
Humanity's Last Exam
A large set of very hard, expert-written questions across many subjects, built to stay difficult for top AI models.
30.1%
Place 13 of 41
ARC-AGI-2
Visual puzzles where the AI must work out a new rule from a few examples and apply it. They test picking up a new idea quickly, not recalling facts.
61.3%
Place 28 of 86
tied with 1 other
SimpleQA Verified
Short factual questions that each have one correct answer. It measures how often a model gets facts right.
23.8%
Place 67 of 82

What it costs in everyday terms

List prices turned into amounts you can picture. Real bills depend on how you use it.

100 chats: $0.04; Cost of 100 chats assumes each chat sends 1,000 tokens (about 750 words) and gets 500 tokens (about 375 words) back.
A 300-page book: $0.01; Cost to read a 300-page book assumes about 100,000 input tokens and no reply. A real book varies with page size and language.
Blended price: $0.20 per 1M; Three parts input to one part output. Our own calculation for comparing models.

Similar models

Models from other makers with a similar overall score.

Learn more

Questions about Claude Haiku 5.5

How much does Claude Haiku 5.5 cost?: Claude Haiku 5.5 costs $0.10 per million input tokens and $0.50 per million output tokens, using the list price from Anthropic as of 26 Oct 2026. That is about $0.04 for 100 chats of about 1,000 tokens in and 500 tokens back. Confirm on the maker's site before you buy, because prices change.
What is the context window of Claude Haiku 5.5?: Claude Haiku 5.5 has a context window of 1M tokens, roughly 750,000 words of text read at once. A single answer can be up to 128K tokens long.
Is Claude Haiku 5.5 open source?: We list Claude Haiku 5.5 as closed: it is used through its maker's apps or API and its weights are not published (Epoch AI lists its access as "API access").
How good is Claude Haiku 5.5 compared with other AI models?: On the Epoch Capabilities Index, Claude Haiku 5.5 scores 159.3 (likely range 156.2 to 162.5), which ranks 16 of 270 models we track. On Arena's human-vote text board it is ranked 87 of 166. Scores close together are a tie in practice.
How does Claude Haiku 5.5 compare with GPT-5.4 Pro?: GPT-5.4 Pro scores higher on 5 of 6 shared quality measures, Claude Haiku 5.5 on 0. 1 measure is too close to call. Claude Haiku 5.5 costs about 100% less than GPT-5.4 Pro on a blended price of $0.20 against $67.50 per million tokens. See the full comparison for every measure.
How current are these numbers for Claude Haiku 5.5?: The newest data on this page is dated 26 Oct 2026. Scores come from Epoch AI and Arena, prices from the open LiteLLM price list.

Where this data comes from

Epoch AI: Benchmark data: Epoch AI, 'Capabilities & benchmarking', epoch.ai/benchmarks (CC BY 4.0). We match models across sources, rank them and show them in charts. Scores are not changed. Epoch AI does not endorse this site.
Arena (formerly LMArena): Human-preference ratings: Arena (arena.ai) leaderboard dataset, CC BY 4.0. We keep the latest published ratings, match them to models from other sources and rank them. Ratings are not changed. Arena does not endorse this site.
LiteLLM model price list: Prices and context windows: LiteLLM model price list (github.com/BerriAI/litellm, MIT licence). We use the model maker's own price entry when there is one. Prices change often, so always confirm on the maker's site before you buy.

Newest data on this page: 26 Oct 2026. Prices are list prices and change often, so confirm on the maker's site before you buy. Sponsors never change a score or an order. How we calculate this · Report a mistake