Anthropic Claude Haiku 5.5 API Pricing and Speed
Anthropic Claude Haiku 5.5 API pricing: $0.10 per 1M tokens for input and $0.50 per 1M tokens for output, from the official API price list of Anthropic. Official rate checked (source); OpenRouter is a separate channel below. Our measured latency and output speed are shown by route and country below. Published OpenRouter: Last checked · Source (data) · Methodology
OpenRouter: Amount worked out from the source page: $0.10 per 1M tokens, in USD · as shown in the United States on 9 Nov 2026 · the page does not say whether tax is included; we do not assume it is excluded · no price change seen since 16 Oct 2026.
Quality on your task: not measured by Measured Lab yet. On our data, Claude Haiku 5.5 fits 6 of 6 common tasks by published prices, context window and terms, with calculated monthly costs.
What can I use it for?
Six common tasks cost a month at list prices
- Calculated Answer customer messages: $2.12 for 5,000 customer messages a month
- Calculated Pull data from invoices and documents with strict JSON and page images: $0.80 for 2,000 documents a month
- Calculated Summarize documents and meetings: $0.98 for 1,000 documents or meetings a month
- Calculated Translate content: $1.65 for 3,000 pages a month
- Calculated Chat assistant on your website: $20.85 for 10,000 conversations a month
- Calculated Write and review code with tool calls: $5.88 for 500 coding tasks a month
What our data confirms
- Observed Strict JSON output: accepted in our test call on 17 Oct 2026
- Published Calls your tools, in Anthropic documentation, read 17 Oct 2026
- Published Reads images; OpenRouter listing, 9 Nov 2026
Example to try
“Review this pull request and list bugs and missing tests, with the file and line of each.” Run it on your own data before you choose.
What we measured
- Observed Speed on a short test request: first token 0.5 s, 221 tokens/s (median of 115 test requests on 23 dates, short request)
Stated by the provider
- Stated by the provider Anthropic for this model: SWE-bench Pro 64.8%, SWE-bench Multilingual 83.7%, Terminal-Bench 4.0 39.2%; quotes and test conditions in Quality signals
Before you choose
- Published Training on your data, per Anthropic Commercial Terms of Service: no, read 9 Nov 2026
- Unknown Where requests are processed: Unknown
Prices USD per 1M tokens, as listed
- $0.10Input per 1M tokens, from Anthropic Published
- $0.50Output per 1M tokens, from Anthropic Published
- $0.01Cached input per 1M tokens, from Anthropic Published
All prices by channel
| Price per 1M tokens | Anthropic direct | Through OpenRouter | Previous prices by channel |
|---|---|---|---|
| Input | $0.10 Published | $0.10 Published | — |
| Output | $0.50 Published | $0.50 Published | — |
| Cached input · read | $0.01 Published | $0.01 Published | — |
| Cache write · 5-minute lifetime | $0.125 Published | — | — |
| Cache write · 1-hour lifetime | $0.20 Published | — | — |
| Cache write | UnknownUnknown | $0.125 Published | — |
| Above the long-request threshold · input | $0.50 Published | — | — |
| Above the long-request threshold · output | $2.50 Published | — | — |
| Above the long-request threshold · cached input | $0.05 Published | — | — |
| Batch · input / output | — | $0.05 / $0.25 Published | — |
Input price from Anthropic, 19 Oct 2026 – 9 Nov 2026
19 Oct 2026$0.10 throughout9 Nov 2026
Input price through OpenRouter, 16 Oct 2026 – 9 Nov 2026
16 Oct 2026$0.10 throughout9 Nov 2026
Context window1,000,000tokens Published
Max output128,000tokens Published
Long-request price above100,000tokens Published
OpenRouter providers who serves this model through OpenRouter, at what price
OpenRouter sends a call without a pinned provider to one of these endpoints, and each endpoint has its own price. Prices are USD per 1M tokens, input / output.
Pinned min: $0.10 / $0.50 · Azure (azure/global)
Unpinned range: input $0.10–$0.11, output $0.50–$0.55 across 10 endpoints of 5 providers (listing shows $0.10 / $0.50).
Models with the widest provider price spread on OpenRouter
Request fragment that pins Azure (azure/global):
{"provider": {"order": ["azure/global"], "allow_fallbacks": false}}With fallbacks turned off, OpenRouter does not move the call to another provider. A full endpoint tag pins that endpoint; the provider name alone covers all of its endpoints (OpenRouter provider routing).
All endpoints (10), by input price
| Provider | Endpoint tag | Precision | Input | Output | Context, tokens |
|---|---|---|---|---|---|
| Amazon Bedrock | amazon-bedrock | Not stated | $0.10 | $0.50 | 1,000,000 |
| Anthropic | anthropic | Not stated | $0.10 | $0.50 | 1,000,000 |
| Azure | azure/global | Not stated | $0.10 | $0.50 | 1,000,000 |
| Claude Platform on AWS | claude-on-aws | Not stated | $0.10 | $0.50 | 1,000,000 |
google-vertex/global | Not stated | $0.10 | $0.50 | 1,000,000 | |
| Amazon Bedrock | amazon-bedrock/eu-west-1 | Not stated | $0.11 | $0.55 | 1,000,000 |
| Amazon Bedrock | amazon-bedrock/us-east-1 | Not stated | $0.11 | $0.55 | 1,000,000 |
| Azure | azure/us | Not stated | $0.11 | $0.55 | 1,000,000 |
google-vertex/europe | Not stated | $0.11 | $0.55 | 1,000,000 | |
google-vertex/us | Not stated | $0.11 | $0.55 | 1,000,000 |
Sorted by input price, then output price. Precision is the quantization field of the endpoints list.
Measured by us our own short test request: about 380 tokens in, a short JSON answer out · medians from answers on two dates or more
United States
| Route | First answer token | Output speed | Cost per request | Hidden reasoning |
|---|---|---|---|---|
| Direct API · Anthropicreasoning: off115 answers on 23 dates · 100 % answered | 528 msrange 0.29–7.55 s · slow 10 %: 1.14 s Observed | 221 tokens/srange 87.6–430 · slowest 10 %: 155 tokens/s | $0.000081 Calculated | 0for 64 visible |
United Kingdom
| Route | First answer token | Output speed | Cost per request | Hidden reasoning |
|---|---|---|---|---|
| Direct API · Anthropicreasoning: off105 answers on 21 dates · 100 % answered | 671 msrange 0.46–7.01 s · slow 10 %: 1.26 s Observed | 243 tokens/srange 56.8–400 · slowest 10 %: 170 tokens/s | $0.000081 Calculated | 0for 64 visible |
Germany
| Route | First answer token | Output speed | Cost per request | Hidden reasoning |
|---|---|---|---|---|
| Direct API · Anthropicreasoning: off105 answers on 21 dates · 100 % answered | 653 msrange 0.49–3.42 s · slow 10 %: 1.05 s Observed | 215 tokens/srange 130–414 · slowest 10 %: 165 tokens/s | $0.000081 Calculated | 0for 64 visible |
Canada
| Route | First answer token | Output speed | Cost per request | Hidden reasoning |
|---|---|---|---|---|
| Direct API · Anthropicreasoning: off105 answers on 21 dates · 100 % answered | 562 msrange 0.33–1.75 s · slow 10 %: 1.02 s Observed | 231 tokens/srange 101–414 · slowest 10 %: 168 tokens/s | $0.000081 Calculated | 0for 64 visible |
Australia
| Route | First answer token | Output speed | Cost per request | Hidden reasoning |
|---|---|---|---|---|
| Direct API · Anthropicreasoning: off105 answers on 21 dates · 100 % answered | 763 msrange 0.58–3.14 s · slow 10 %: 1.15 s Observed | 227 tokens/srange 120–387 · slowest 10 %: 164 tokens/s | $0.000081 Calculated | 0for 64 visible |
- Hidden reasoning tokens are billed as output, at the output price, even though they are not part of the answer.
How the test works and what each number means: methodology. The figures are also in our open data.
Quality signals public test results, each with its source and date
| Signal | Result | Source |
|---|---|---|
| Instruction followingIFEval, 50 prompts · reasoning: off | 82 % Observedof 50 prompts with every instruction followed | Our run of a fixed IFEval subset, checked by code82 % on 22 Oct 2026 · 88 % on 17 Oct 2026 · instruction level 86 % |
Figures the provider publishes (7)
| Benchmark | Provider's figure | Where the provider says it |
|---|---|---|
| SWE-bench Pro | 64.8% Stated by the provider | Claude Haiku 5.5 achieved 64.8%.Claude Haiku 5.5 System Card (October 7, 2026)read 17 Oct 2026 |
| SWE-bench Multilingual | 83.7% Stated by the provider | Haiku 5.5 achieved 83.7%.Claude Haiku 5.5 System Card (October 7, 2026)read 17 Oct 2026 |
| Terminal-Bench 4.0 | 39.2% Stated by the provider | Claude Haiku 5.5 scored 39.2% on Terminal-Bench 4.0 with safeguards enabled.Claude Haiku 5.5 System Card (October 7, 2026)read 17 Oct 2026 |
| Humanity's Last Examno tools | 45.9% Stated by the provider | Results table in Introducing Claude Haiku 5.5read 17 Oct 2026 |
| Humanity's Last Examwith tools | 57.4% Stated by the provider | Results table in Introducing Claude Haiku 5.5read 17 Oct 2026 |
| OSWorld 2.1 (offline subset) | 72.4% Stated by the provider | Claude Haiku 5.5 achieved a partial score of 72.4% and a strict pass rate of 37.1%.Claude Haiku 5.5 System Card (October 7, 2026)read 17 Oct 2026 |
| GDPval-AA v2.1at max effort | 1620 Stated by the provider | Claude Haiku 5.5 scored an Elo of 1620 at max effort, up from 735 for Claude Haiku 4.5.Claude Haiku 5.5 System Card (October 7, 2026)read 17 Oct 2026 |
- Your own task: quality on it is not measured here. These are public test sets; a short trial on a few of your own examples is the check that fits your work.
- Other public leaderboards, without figures copied here: human preference votes on open-ended answers: Arena (LMArena) text leaderboard; long tasks in coding, data analysis and research: LiveBench.
What each signal covers and how our run is scored: methodology.
How we got this number
- What we read: the published price on openrouter.ai/api/v1/models
- Checked: ; tracked since 16 Oct 2026
- Shown in: the United States
- Checks so far: 25 checks since 16 Oct 2026
- What changed: no price change since 16 Oct 2026
- Page updated:
Other measured facts
| Question | Answer | Checked | Status |
|---|---|---|---|
| Above how many request tokens does Claude Haiku 5.5 switch to its long-request price directly from Anthropic (official API price)? | 100,000 tokens | Published | |
| How much does Anthropic Claude Haiku 5.5 cost per 1M cache-read tokens (cache hits) on OpenRouter? | $0.01 per 1M tokens | Published | |
| How much does Anthropic Claude Haiku 5.5 cost per 1M cache-write tokens on OpenRouter? | $0.13 per 1M tokens | Published | |
| How much does Anthropic Claude Haiku 5.5 cost per 1M output tokens on OpenRouter? | $0.50 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M cache-read tokens (cache hits) directly from Anthropic (official API price)? | $0.01 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M cache-read tokens (cache hits) when a request is above the long-request threshold, directly from Anthropic (official API price)? | $0.05 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M cache-write tokens when a request is above the long-request threshold, directly from Anthropic (official API price)? | $0.63 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M cache-write tokens with a 1-hour cache lifetime directly from Anthropic (official API price)? | $0.20 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M cache-write tokens with a 5-minute cache lifetime directly from Anthropic (official API price)? | $0.13 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M input tokens directly from Anthropic (official API price)? | $0.10 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M output tokens directly from Anthropic (official API price)? | $0.50 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M request tokens when a request is above the long-request threshold, directly from Anthropic (official API price)? | $0.50 per 1M tokens | Published | |
| How much does Claude Haiku 5.5 cost per 1M response tokens when a request is above the long-request threshold, directly from Anthropic (official API price)? | $2.50 per 1M tokens | Published | |
| What is the context window of Anthropic Claude Haiku 5.5 on OpenRouter, in tokens? | 1,000,000 tokens | Published | |
| What is the maximum output length of Anthropic Claude Haiku 5.5 on OpenRouter, in tokens? | 128,000 tokens | Published |
History
Same price in all 25 checks since 16 Oct 2026.
- — First check: $0.10 per 1M tokens
The source record lists $0.0000001 per token ($0.10 per 1M tokens), fetched on 9 Nov 2026.
The source record, as published
"prompt":"0.0000001"
Latest briefings each price change recalculated as a task cost
- DeepSeek V4 Pro 0813: price rise input $0.115 → $0.111 per 1M tokens (-3 %), output $4.00 → $4.44 per 1M tokens (+11 %)
- DeepSeek V4.1 Flash 1017: price cut input $0.300 → $0.0060 per 1M tokens (-98 %), output $1.20 → $0.500 per 1M tokens (-58 %)
- MoonshotAI Kimi K2.5: price rise input $0.500 → $0.532 per 1M tokens (+6 %), output $2.50 → $3.33 per 1M tokens (+33 %)
Record history — 346 records for 16 questions
- Last record
- Shown in
- the United States
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.10 per 1M tokens | Published |
| 8 Nov 2026 | $0.10 per 1M tokens | Published |
| 7 Nov 2026 | $0.10 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | 100,000 tokens | Published |
| 8 Nov 2026 | 100,000 tokens | Published |
| 7 Nov 2026 | 100,000 tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.01 per 1M tokens | Published |
| 8 Nov 2026 | $0.01 per 1M tokens | Published |
| 7 Nov 2026 | $0.01 per 1M tokens | Published |
| 6 Nov 2026 | $0.01 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.13 per 1M tokens | Published |
| 8 Nov 2026 | $0.13 per 1M tokens | Published |
| 7 Nov 2026 | $0.13 per 1M tokens | Published |
| 6 Nov 2026 | $0.13 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.50 per 1M tokens | Published |
| 8 Nov 2026 | $0.50 per 1M tokens | Published |
| 7 Nov 2026 | $0.50 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.01 per 1M tokens | Published |
| 8 Nov 2026 | $0.01 per 1M tokens | Published |
| 7 Nov 2026 | $0.01 per 1M tokens | Published |
| 6 Nov 2026 | $0.01 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.05 per 1M tokens | Published |
| 8 Nov 2026 | $0.05 per 1M tokens | Published |
| 7 Nov 2026 | $0.05 per 1M tokens | Published |
| 6 Nov 2026 | $0.05 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.63 per 1M tokens | Published |
| 8 Nov 2026 | $0.63 per 1M tokens | Published |
| 7 Nov 2026 | $0.63 per 1M tokens | Published |
| 6 Nov 2026 | $0.63 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.20 per 1M tokens | Published |
| 8 Nov 2026 | $0.20 per 1M tokens | Published |
| 7 Nov 2026 | $0.20 per 1M tokens | Published |
| 6 Nov 2026 | $0.20 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.13 per 1M tokens | Published |
| 8 Nov 2026 | $0.13 per 1M tokens | Published |
| 7 Nov 2026 | $0.13 per 1M tokens | Published |
| 6 Nov 2026 | $0.13 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.10 per 1M tokens | Published |
| 8 Nov 2026 | $0.10 per 1M tokens | Published |
| 7 Nov 2026 | $0.10 per 1M tokens | Published |
| 6 Nov 2026 | $0.10 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.50 per 1M tokens | Published |
| 8 Nov 2026 | $0.50 per 1M tokens | Published |
| 7 Nov 2026 | $0.50 per 1M tokens | Published |
| 6 Nov 2026 | $0.50 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $0.50 per 1M tokens | Published |
| 8 Nov 2026 | $0.50 per 1M tokens | Published |
| 7 Nov 2026 | $0.50 per 1M tokens | Published |
| 6 Nov 2026 | $0.50 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | $2.50 per 1M tokens | Published |
| 8 Nov 2026 | $2.50 per 1M tokens | Published |
| 7 Nov 2026 | $2.50 per 1M tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | 1,000,000 tokens | Published |
| 8 Nov 2026 | 1,000,000 tokens | Published |
| 7 Nov 2026 | 1,000,000 tokens | Published |
| 6 Nov 2026 | 1,000,000 tokens | Published |
| Time | Value | Status |
|---|---|---|
| 9 Nov 2026 | 128,000 tokens | Published |
| 8 Nov 2026 | 128,000 tokens | Published |
| 7 Nov 2026 | 128,000 tokens | Published |
| 6 Nov 2026 | 128,000 tokens | Published |
Showing the latest 60 of 346 records; the full history is in data.json and data.csv.