What is the cheapest Anthropic model to use via API?

Claude Haiku 4.5 is the most cost-effective option, with input tokens priced at $1 per million and output tokens at $5 per million.

How much can be saved by switching to annual billing?

Annual billing reduces the cost of a Team Standard seat from $25 to $20 per month and a Team Premium seat from $125 to $100 per month.

What are the current rates for Claude 3.5 Sonnet?

For Claude 3.5 Sonnet, batch input tokens are $3 per million and batch output tokens are $15 per million. Public extended access rates are higher at $6 input and $30 output per million tokens.

Does Anthropic offer an Enterprise plan?

Yes, Claude Enterprise is available at $20 per seat per month when billed annually.

ai-costs

Anthropic API Pricing History

By Paige Gilmore, Founder, NetLift· Published July 28, 2026· Updated July 28, 2026

Anthropic API pricing has remained stable through July 2026, with model usage starting at $1 per million input tokens for Haiku 4.5. Seat-based plans for teams range from $20 to $125 per month depending on the tier and billing cycle.

14-day free trial. No credit card required.

Anthropic API pricing currently follows a tiered structure based on model capability and user access. The entry-level Haiku 4.5 model is priced at $1 per million input tokens, while high-performance models like Fable 5 reach $10 per million input tokens. Monthly seat-based subscriptions for team collaboration remain unchanged in recent records.

Finance and AI leaders can choose between per-token API consumption or per-seat subscriptions. Annual commitments offer a 20% discount on seat-based plans, such as the Team Standard tier which reduces from $25 to $20 per seat per month.

Anthropic API pricing history

Date	Plan	Previous price	New price
2026-07-28	Claude — Team - Standard seat (Monthly)	$25	$25
2026-07-28	Claude — Team - Standard seat (Annual)	$20	$20
2026-07-28	Claude — Team - Premium seat (Monthly)	$125	$125
2026-07-28	Claude — Team - Premium seat (Annual)	$100	$100
2026-07-28	Claude — Pro (Monthly)	$20	$20
2026-07-28	Claude — Pro (Monthly)	$20	$20
2026-07-28	Claude — Pro (Annual)	$17	$17
2026-07-28	Claude — Pro (Annual)	$17	$17
2026-07-28	Claude — Max	$100	$100
2026-07-28	Claude — Max	$100	$100
2026-07-28	Claude — Free	$0	$0
2026-07-28	Claude — Free	$0	$0
2026-07-28	Claude — Enterprise	$20	$20
2026-07-28	Claude — Enterprise	$20	$20
2026-07-28	Claude Platform — Web search	$10	$10
2026-07-28	Claude Platform — Managed Agents	$0.08	$0.08
2026-07-28	Claude Platform — Code execution	$0.05	$0.05
2026-07-28	Claude Platform Features — Web search	$10	$10
2026-07-28	Claude Platform Features — Managed Agents	$0.08	$0.08
2026-07-28	Claude Platform Features — Code execution	$0.05	$0.05

NetLift has monitored the official Anthropic API pricing page since 2026-07-27. 20 change(s) recorded so far.

Current verified prices

Plan	Price	Billing	Notes
Claude — Pro (Annual)	$17	per month	$200 billed up front for the year.
Claude — Pro (Monthly)	$20	per month	Standard monthly billing.
Claude — Team Standard (Annual)	$20	per seat / month	For teams of 2 to 150. Billed annually.
Claude — Team Standard (Monthly)	$25	per seat / month	For teams of 2 to 150.
Claude — Max	$100	per month	Pricing starts from $100 per month.
Claude — Team Premium (Annual)	$100	per seat / month	Billed annually.
Claude — Team Premium (Monthly)	$125	per seat / month
Claude Enterprise — Enterprise	$20	per seat / month	Billed annually. Sales-assisted or self-serve options available.
Claude Team — Standard seat (Annual)	$20	per seat / month	Billed annually
Claude Team — Standard seat (Monthly)	$25	per seat / month
Claude Team — Premium seat (Annual)	$100	per seat / month	Billed annually
Claude Team — Premium seat (Monthly)	$125	per seat / month

Current verified per-token prices

Model / rate	Price	Unit
Anthropic Claude 3.5 Sonnet — Batch Input tokens	$3	per 1M input tokens
Anthropic Claude 3.5 Sonnet — Public Extended Access Input tokens	$6	per 1M input tokens
Anthropic Claude 3.5 Sonnet — Batch Output tokens	$15	per 1M output tokens
Anthropic Claude 3.5 Sonnet — Public Extended Access Output tokens	$30	per 1M output tokens
Anthropic Claude Opus — On-Demand Input tokens	$6	per 1M input tokens
Anthropic Claude Opus — On-Demand Output tokens	$30	per 1M output tokens
Anthropic Claude Sonnet 5 — Promotional Input tokens	$2	per 1M input tokens
Anthropic Claude Sonnet 5 — Promotional Output tokens	$10	per 1M output tokens
Claude API — Opus 4.8 - Input	$5	per 1M input tokens
Claude API (Batch Legacy) — Sonnet 4.6 Input	$3	per 1M input tokens
Claude API (Batch Legacy) — Opus 4.8 Input	$5	per 1M input tokens
Claude API (Batch) — Haiku 4.5 Input	$1	per 1M input tokens
Claude API (Batch) — Sonnet 5 Input	$2	per 1M input tokens
Claude API (Batch) — Opus 5 Input	$5	per 1M input tokens
Claude API (Batch) — Haiku 4.5 Output	$5	per 1M output tokens
Claude API (Batch) — Fable 5 Input	$10	per 1M input tokens
Claude API (Batch) — Sonnet 5 Output	$10	per 1M output tokens
Claude API (Batch) — Opus 5 Output	$25	per 1M output tokens
Claude API (Batch) — Fable 5 Output	$50	per 1M output tokens
Claude API (Fable 5) — Prompt caching - Read	$1	per 1M input tokens
Claude API (Fable 5) — Input	$10	per 1M input tokens
Claude API (Fable 5) — Prompt caching - Write	$12.50	per 1M input tokens
Claude API (Fable 5) — Output	$50	per 1M output tokens
Claude API (Haiku 4.5) — Prompt caching - Read	$0.1	per 1M input tokens
Claude API (Haiku 4.5) — Input	$1	per 1M input tokens
Claude API (Haiku 4.5) — Prompt caching - Write	$1.25	per 1M input tokens
Claude API (Haiku 4.5) — Output	$5	per 1M output tokens
Claude API (Opus 4.1) — Input	$15	per 1M input tokens
Claude API (Opus 4.1) — Output	$75	per 1M output tokens
Claude API (Opus 5) — Prompt caching - Read	$0.5	per 1M input tokens
Claude API (Opus 5) — Input	$5	per 1M input tokens
Claude API (Opus 5) — Prompt caching - Write	$6.25	per 1M input tokens
Claude API (Opus 5) — Output	$25	per 1M output tokens
Claude API (Sonnet 4.6) — Input	$3	per 1M input tokens
Claude API (Sonnet 4.6) — Output	$15	per 1M output tokens
Claude API (Sonnet 5) — Prompt caching - Read	$0.2	per 1M input tokens
Claude API (Sonnet 5) — Input	$2	per 1M input tokens
Claude API (Sonnet 5) — Prompt caching - Write	$2.50	per 1M input tokens
Claude API (Sonnet 5) — Output	$10	per 1M output tokens
Claude Platform — Prompt Caching (Read) - Opus 5	$0.5	per 1M input tokens

14-day free trial. No credit card required.

How do token costs scale across models?

Usage costs are determined by the specific model version selected for the API call. Input rates range from $1 per million tokens on Haiku 4.5 to $10 per million on Fable 5. Output tokens are consistently priced at 5x the input rate for most current models. Organizations using batch processing for non-urgent tasks can access lower rates, such as $3 per million input tokens for Claude 3.5 Sonnet.

What are the costs for platform features?

Beyond raw token consumption, Anthropic charges for specific platform capabilities. Web search functionality is priced at $10, while managed agents and code execution cost $0.08 and $0.05 respectively. Prompt caching is also available to reduce costs on repetitive inputs, with read rates as low as $0.10 per million tokens on the Haiku 4.5 model.

Measuring the return on Anthropic investment requires comparing API consumption against the actual labor hours recovered in production workflows. NetLift helps organizations move beyond monitoring credit spend to quantifying the net return by mapping model costs to specific business outcomes and time-saved baselines.

14-day free trial. No credit card required.

Sources

All pricing on this page comes from official vendor pages:

https://www.anthropic.com/pricing — retrieved 2026-07-28
https://claude.com/pricing — retrieved 2026-07-27
https://aws.amazon.com/bedrock/pricing/ — retrieved 2026-07-28

Frequently Asked Questions

Ready to see your AI return?

14-day free trial. No credit card required.

About the author

Paige Gilmore · Founder, NetLift

Paige Gilmore is the founder of NetLift, the AI Value Management platform that helps organisations measure the cost, savings and return of AI adoption.

Paige Gilmore on LinkedIn

Keep reading

ai-costs

Gemini API Pricing History

Gemini API pricing has remained stable since monitoring began in July 2026, with the Gemini 3.5 Live Translate output rate holding at $21 per 1 million tokens. Current costs range from $0.25 per 1 million input tokens for Flash-Lite models up to $21 for specialized translation outputs.

Read ai-costs

OpenAI API Pricing History

On July 27, 2026, OpenAI moved its ChatGPT Go, Plus, and Pro plans from fixed monthly rates to a sales-led model. Current API rates vary widely by model performance, with input costs ranging from $0.20 to $32.00 per 1M tokens.

Read ai-costs

Gemini Pricing History

Gemini pricing has remained stable since monitoring began in July 2026, with costs ranging from $0.25 to $21 per million tokens. The pricing structure favors high-volume efficiency through tiered rates for context length and processing priority.

Read ai-costs

GitHub Copilot Pricing History

GitHub Copilot pricing currently ranges from $0 to $100 per seat per month. Since tracking began in July 2026, costs for the Pro, Pro+, and Max plans have remained stable.

Read ai-costs

Microsoft Copilot Pricing History

Microsoft Copilot pricing ranges from $18 to $38.40 per seat per month, with the standard enterprise add-on holding steady at $30 per month for annual commitments. Pricing for core plans has remained stable since July 2026, though some business add-ons have seen price reductions.

Read