Claude API Pricing
The AI safety-focused lab behind Claude. Founded by ex-OpenAI researchers in 2021, it ships frontier-grade reasoning models through an API that does not retain prompts by default, and publishes a Responsible Scaling Policy, now in version 3.0.
Claude Sonnet 5.5 LAUNCHED SEP 2026
Claude Sonnet 5.5 replaced Claude Sonnet 5 as Anthropic's current Sonnet on 2026-09-28 at Sonnet 5's rate card: $2/M input, $10/M output, $2.50/M 5-minute and $4/M 1-hour cache writes, batch half p…
Claude Opus 5.5 LAUNCHED SEP 2026
Claude Opus 5.5 replaced Claude Opus 5 as Anthropic's recommended default on 2026-09-22 and costs less than the model it replaces: $4/M input and $20/M output against Opus 5's $5/$25.
Claude Haiku 5.5 LAUNCHED OCT 2026
Claude Haiku 5.5 is Anthropic's current Haiku, released 2026-10-07 for high-volume, latency-sensitive work such as classification, routing, extraction and subagent tasks.
Claude Fable 5.1 LAUNCHED SEP 2026
Claude Fable 5.1 shipped 2026-09-01 alongside its sibling.
Claude Mythos 5.1 LAUNCHED SEP 2026
⚠ Glasswing programme only - vetted cybersecurity and life-sciences organisations. No self-serve API alias published. Claude Mythos 5.1 shipped 2026-09-01 alongside its sibling.
Claude Sonnet 5
Claude Sonnet 5 is the Sonnet generation before 5.5, which replaced it as Anthropic's current Sonnet on 2026-09-28 at the same prices; the models overview now lists Sonnet 5 among legacy models tha…
Claude Opus 5 LAUNCHED JUL 2026
Claude Opus 5 is the Opus generation before 5.5, which replaced it as Anthropic's recommended default on 2026-09-22.
Claude Fable 5 LAUNCHED JUN 2026
Claude Fable 5 is listed by Anthropic at $10/M input, $12.50/M 5m cache writes, $20/M 1h cache writes, $1/M cache hits, and $50/M output.
Claude Mythos 5 LAUNCHED JUN 2026
⚠ Glasswing partners first; broader trusted access planned. No self-serve API alias published in launch post. Claude Mythos 5 is listed by Anthropic at $10/M input, $12.50/M 5m cache writes, $20/M …
Claude Opus 4.8 LAUNCHED MAY 2026
Newest generally available Opus release.
Claude Opus 4.7
Newest Opus.
Claude Opus 4.6
.
Claude Sonnet 4.6
.
Claude Opus 4.5
Launched 2025-11-24 at $5/M input and $25/M output - the announcement reads 'Pricing is now $5/$25 per million tokens', a third of the $15/$75 of Opus 4.1 and earlier Opus models; unchanged since.
Claude Haiku 4.5
.
Claude Sonnet 4.5
Claude Sonnet 4.5 (claude-sonnet-4-5-20250929), released 2025-09-29 at the Sonnet price band: $3/M input, $3.75/M 5-minute cache writes, $6/M 1-hour cache writes, $0.30/M cache hits, $15/M output, …
Claude Opus 4.1
Legacy Opus pricing tier at $15/M input, $1.50/M cache-hit, $75/M output.
Claude Opus 4 LAUNCHED MAY 2025
Deprecated on April 14, 2026 with retirement scheduled for June 15, 2026.
Claude Sonnet 4 LAUNCHED MAY 2025
API model name claude-sonnet-4-20250514.
Claude Sonnet 3.7
Archive-only model.
Claude Sonnet 3.5 (October 2024) LAUNCHED OCT 2024
API model name claude-3-5-sonnet-20241022.
Claude Haiku 3.5
Retired from first-party Claude API on Feb 19, 2026 (still available on Amazon Bedrock and Vertex AI per their own deprecation schedules).
Claude Sonnet 3.5 (June 2024) LAUNCHED JUN 2024
API model name claude-3-5-sonnet-20240620.
Claude Opus 3 LAUNCHED MAR 2024
API model name claude-3-opus-20240229.
Claude Sonnet 3 LAUNCHED MAR 2024
API model name claude-3-sonnet-20240229.
Claude Haiku 3 LAUNCHED MAR 2024
API model name claude-3-haiku-20240307.
Claude Mythos Preview
⚠ Glasswing partners only — no public API. Restricted-access Mythos Preview archive.
| Model | Input /M | Output /M | Cached | Context | Max output | Vision | Tools | Tier |
|---|---|---|---|---|---|---|---|---|
| Claude Sonnet 5.5 | $2.00 | $10.00 | $0.10−95% | 1M | — | ✓ | ✓ | Active |
| Claude Opus 5.5 | $4.00 | $20.00 | $0.20−95% | 1M | — | ✓ | ✓ | Frontier |
| Claude Haiku 5.5 | $0.10 | $0.50 | $0.01−90% | 1M | — | ✓ | ✓ | Light |
| Claude Fable 5.1 FLAGSHIP | $10.00 | $50.00 | $0.25−98% | 1M | — | ✓ | ✓ | Active |
| Claude Mythos 5.1 RESTRICTED | $10.00 | $50.00 | $0.25−98% | 1M | Glasswing programme only - vetted cybersecurity and life-sciences organisations. No self-serve API alias published. | ✓ | ✗ | Restricted |
| Claude Sonnet 5 | $2.00 | $10.00 | $0.20−90% | 1M | — | ✓ | ✓ | Active |
| Claude Opus 5 | $5.00 | $25.00 | $0.50−90% | 1M | — | ✓ | ✓ | Frontier |
| Claude Fable 5 | $10.00 | $50.00 | $1.00−90% | 1M | — | ✓ | ✓ | Active |
| Claude Mythos 5 RESTRICTED | $10.00 | $50.00 | $1.00−90% | 1M | Glasswing partners first; broader trusted access planned. No self-serve API alias published in launch post. | ✓ | ✗ | Restricted |
| Claude Opus 4.8 | $5.00 | $25.00 | $0.50−90% | 1M | — | ✓ | ✓ | Frontier |
| Claude Opus 4.7 | $5.00 | $25.00 | $0.50−90% | 1M | — | ✓ | ✓ | Frontier |
| Claude Opus 4.6 | $5.00 | $25.00 | $0.50−90% | 1M | — | ✓ | ✓ | Frontier |
| Claude Sonnet 4.6 | $3.00 | $15.00 | $0.30−90% | 1M | — | ✓ | ✓ | Mid |
| Claude Opus 4.5 | $5.00 | $25.00 | $0.50−90% | 200K | — | ✓ | ✓ | Frontier |
| Claude Haiku 4.5 | $1.00 | $5.00 | $0.10−90% | 200K | — | ✓ | ✓ | Light |
| Claude Sonnet 4.5 DEPRECATED | $3.00 | $15.00 | $0.30−90% | 200K | — | ✓ | ✓ | Deprecated |
| Claude Opus 4.1 RETIRED | — | — | — | Retired Aug 5, 2026 | → Claude Opus 4.8 | ✓ | ✗ | Retired |
| Claude Opus 4 RETIRED | — | — | — | Retired Jun 15, 2026 | → Claude Opus 4.8 | ✓ | ✗ | Retired |
| Claude Sonnet 4 RETIRED | — | — | — | Retired Jun 15, 2026 | → Claude Sonnet 4.6 | ✓ | ✗ | Retired |
| Claude Sonnet 3.7 RETIRED | — | — | — | Retired Feb 19, 2026 | → Claude Sonnet 4.6 | ✓ | ✗ | Retired |
| Claude Sonnet 3.5 (October 2024) RETIRED | — | — | — | Retired Oct 28, 2025 | → Claude Sonnet 4.6 | ✓ | ✗ | Retired |
| Claude Haiku 3.5 RETIRED | — | — | — | Retired Feb 19, 2026 | → Claude Haiku 4.5 | ✗ | ✗ | Retired |
| Claude Sonnet 3.5 (June 2024) RETIRED | — | — | — | Retired Oct 28, 2025 | → Claude Sonnet 4.6 | ✓ | ✗ | Retired |
| Claude Opus 3 RETIRED | — | — | — | Retired Jan 5, 2026 | → Claude Opus 4.7 | ✓ | ✗ | Retired |
| Claude Sonnet 3 RETIRED | — | — | — | Retired Jul 21, 2025 | → Claude Sonnet 4.6 | ✓ | ✗ | Retired |
| Claude Haiku 3 RETIRED | — | — | — | Retired Apr 20, 2026 | → Claude Haiku 4.5 | ✓ | ✗ | Retired |
| Claude Mythos Preview RESTRICTED | — | — | — | Withdrawn | Glasswing partners only — no public API | ✗ | ✗ | Restricted |
Pricing across the lineup.
The last 12 months at Anthropic.
Major releases, pricing moves, and capability updates. Sourced from official Anthropic announcements + our own verification.
platform.claude.com - direct API access, workbench and billing, at the list prices on this page. Global routing by default; US-only inference costs 1.1x on newer models.
CLOUD · AWSClaude on Amazon Bedrock, billed by AWS. Global endpoints, plus regional endpoints that keep data in a region at a 10% premium for Claude 4.5 and later. Anthropic also runs Claude Platform on AWS.
CLOUD · GCPClaude on Google Cloud (Vertex AI). Global, multi-region and regional endpoints; multi-region and regional carry a 10% premium for Claude 4.5 and later.
CONSUMER · WEBPro / Max / Team subscriptions for web, desktop and mobile. Since October 2026 Max and Team include a monthly API credit: $100 on Max 5x, $200 on Max 20x, up to $500 pooled on Team.
AGENT · CLIAnthropic's coding agent. Included in Pro, Max and Team; also runs on a Console key, billed per token at the rates on this page.
Anthropic was founded in 2021 by Dario and Daniela Amodei alongside several former OpenAI researchers. The pitch: build frontier-grade AI with safety and interpretability as first-class concerns, not afterthoughts.
The company is structured as a public-benefit corporation (PBC), which legally requires its directors to weigh a stated public benefit alongside shareholder return. Its Responsible Scaling Policy, which ties model capabilities to required safeguards, reached version 3.0 on February 24, 2026.
Anthropic's best-known technical contributions are Constitutional AI (training models against explicit written principles) and the Model Context Protocol, the open standard for connecting models to tools and data, which Anthropic donated to the Linux Foundation's Agentic AI Foundation in December 2025.
In May 2026 Anthropic raised $65 billion in a Series H at a $965 billion post-money valuation, led by Altimeter Capital, Dragoneer, Greenoaks and Sequoia Capital; its previous round was a $30 billion Series G at $380 billion. Amazon and Google resell Claude through Bedrock and Google Cloud, and Microsoft through Foundry.
The lineup is narrower than OpenAI's or Google's: Anthropic sells no image-generation model and no embedding model of its own - its documentation points to Voyage AI for embeddings - and concentrates on text-and-image-input models for coding, agents and knowledge work. See comparisons.
OpenAI
GPT-6.1 Sol matches Sonnet 5.5's sticker at $2.00/$10.00 per M; GPT-6 Astra tops the range at $10.00/$50.00. OpenAI also sells image-generation and realtime-audio models, which Anthropic does not.
CLOUD GIANTGemini 3.8 Flash carries a 1M-token window at $0.75/$3.75 per M. Google also sells image-generation and native-audio models, which Anthropic does not.
EU CHAMPIONMistral
EU-based, with open weights for much of its range. Mistral Large 4, in public preview, lists at $1.36/$4.18 per M; Large 3, with downloadable weights, at $0.50/$1.50.
COST DISRUPTORDeepSeek
DeepSeek V4 Pro lists at $1.32/$3.96 per M with a 1M-token window - a fraction of Opus 5.5's $4.00/$20.00. China-based.
FRONTIER RIVALSpaceXAI (Grok)
Grok 4.7 lists at $2.00/$6.00 per M below a 200K-token prompt and doubles above it, with a 500K window.
OPEN-WEIGHT CHALLENGERMoonshot (Kimi)
Kimi K3 lists at $3.00/$15.00 per M with a 1M-token window.
Frequently asked.
Practical questions about Anthropic and the Claude lineup.
Q · 01 What is the newest Claude model? +
$0.10/M input and $0.50/M output for prompts up to 100,000 tokens - 90% below the Haiku 4.5 it replaces, and five times higher above 100,000 tokens. The same day Anthropic halved the cache read on Claude Sonnet 5.5 (September 28) to $0.10/M; its input and output stay at $2.00/$10.00. Claude Opus 5.5 (September 22), Anthropic's recommended default, lists at $4.00/$20.00, and the highest-capability model is Claude Fable 5.1 at $10.00/$50.00.Q · 02 Which Claude model should I start with? +
$4.00/$20.00. Claude Sonnet 5.5 costs half that, $2.00/$10.00, and is faster - the pick when latency and volume matter more than the last step of capability. Go down to Haiku 5.5, $0.10/$0.50 on prompts up to 100,000 tokens, for high-volume classification, extraction and subagent work, and up to Fable 5.1 only where Opus 5.5 at higher effort falls short, at 2.5x the Opus rate.Q · 03 Is Anthropic safer than OpenAI? +
Q · 04 Does Anthropic offer EU data residency? +
inference_geo setting accepts only global (the default) and us, and US-only inference costs 1.1x on models released after February 1, 2026. To keep processing in an EU region, use a regional endpoint on Amazon Bedrock or Google Cloud; Anthropic's pricing page says regional and multi-region endpoints carry a 10% premium over global endpoints for Claude 4.5 and later.Q · 05 What's the data retention policy? +
Q · 06 Is there a Claude 5 generation? +
Q · 07 Can I get a discount? +
Q · 08 How does Anthropic compare to other frontier labs? +
$4.00/$20.00 per M, below GPT-6 Astra's $10.00/$50.00 and well above Google's Flash tier. See comparisons.