NEW SONNETPRICE HELDAGENTICTEXT + VISIONPROMPT CACHING
Claude Sonnet 5 API Pricing
Claude Sonnet 5 is Anthropic's newest Sonnet model for coding, agents, computer use, and high-volume professional workflows. Anthropic lists it at $2/M input and $10/M output. That rate launched as introductory pricing due to expire August 31, 2026 — on August 13, 2026 Anthropic made it the standard price and cancelled the scheduled rise to $3/$15.
Input - per 1M tokens
$2.00/M
Standard since Aug 13 held
Output - per 1M tokens
$10.00/M
$15/M rise cancelled held
Cached input - per 1M tokens
$0.20/M
Cache write $2.50 / $4 -90%
Effective - agentic blend
$1.36/M
92/8 split - 82% cache
§ 01 / TERMINAL
Run the numbers.
Calculator pre-loaded with the standard Claude Sonnet 5 API rate. No price change is scheduled: the $3/M input and $15/M output tier once planned for September 1, 2026 was cancelled.
$ /mo
Workload split
Prompt cache hit rate
Tokens you can process
—
Words equivalent (English)
—
Effective rate
—
§ 02 / SCENARIOS
Real-world presets.
SUPPORT AGENT
Ticket triage with tools
$0.022/ticket
CODING AGENT
Repo-wide bug fix
$0.194/task
LONG DOC ANALYSIS
Contract risk review
$0.227/doc
RESEARCH OPS
Analyst brief draft
$0.077/brief
§ 03 / TOKENIZER
Paste text. See tokens. See cost.
Estimate · anthropic-bpe-estimate · ≈2.6 chars/token Auto-counts as you type
This is a chars-per-token approximation, not a real tokenizer. Actual tokens vary by language, code density, and tool-call overhead — counts are typically ±10–20% off for English prose, more for code or non-Latin scripts. For exact billing, use the vendor's official tokenizer.
Characters 635
Words 105
Tokens (estimated) 244 tokens
Cost as input · uncached $0.00049 USD
Cost as output · uncached $0.00244 USD
Cost as cached input $0.00005 USD
| Model | Input /M | Output /M | Effective blended | Context | Best for |
|---|---|---|---|---|---|
| Claude Sonnet 5 Current | $2.00 cache $0.20 | $10.00 | $1.36 standard price | 1M | New Sonnet daily driver |
| Claude Sonnet 4.6 | $3.00 cache $0.30 | $15.00 | $2.05 standard Sonnet price | 1M | Existing Sonnet migrations |
| Claude Opus 5 | $5.00 cache $0.50 | $25.00 | $3.41 harder reasoning sibling | 1M | Anthropic's general frontier tier |
| Claude Haiku 4.5 | $1.00 cache $0.10 | $5.00 | $0.682 cheaper Anthropic tier | 200K | Support and classification |
| Claude Fable 5 | $10.00 cache $1.00 | $50.00 | $6.82 higher-tier Claude | 1M | Anthropic's top capability tier |
| GPT-5.6 Sol | $4.00 cache $0.40 | $20.00 | $2.73 frontier competitor | 1.05M | OpenAI frontier agents |
| Gemini 3.6 Flash | $0.75 cache $0.075 | $3.75 | $0.481 cheaper multimodal peer | 1M | Google's mainstream tier |
| Fugu Ultra | $5.00 cache $0.50 | $30.00 | $3.61 multi-agent competitor | 272K | Orchestrated multi-agent tasks |
Frequently asked.
Practical Claude Sonnet 5 pricing questions, including the standard API rate, the cancelled September 1 increase, and how cache writes are billed.
Q · 01 What is Claude Sonnet 5 priced at? +
Anthropic lists Claude Sonnet 5 at
$2/M input and $10/M output. That began as introductory pricing due to end August 31, 2026, but on August 13, 2026 Anthropic made it the standard price: the page now says the scheduled increase to $3/$15 on September 1, 2026 will not occur.Q · 02 How is cached input billed? +
Anthropic's pricing table itemises Claude Sonnet 5: cache hits at
$0.20/M, 5-minute cache writes at $2.50/M and 1-hour writes at $4/M. Writes replace the base input price rather than adding to it, so a 5-minute cache pays for itself after a single hit. These figures are stable: the September 1, 2026 increase that would have raised them to $0.30/M, $3.75/M and $6/M was cancelled.Q · 03 What model ID should developers use? +
Anthropic says to use
claude-sonnet-5 via the Claude API. It is available natively on Claude Platform and through AWS, Google Cloud, and Microsoft Foundry.Q · 04 Is Sonnet 5 still going up in price on September 1? +
No. Anthropic cancelled it. The launch schedule had Sonnet 5 moving to
$3/M input and $15/M output on September 1, 2026 - a 50% rise that would have taken the blended agentic rate from $1.36/M to about $2.05/M. On August 13, 2026 the pricing page stated the increase will not occur and that $2/$10 is now standard. Budgets built around a September step-up can be unwound.Q · 05 Is there a Batch API discount? +
Yes. Anthropic's Sonnet 5 page points to 50% savings with batch processing, which on the standard rate implies
$1/M input and $5/M output for eligible batch workloads.Q · 06 Does US-only inference change the price? +
Yes. Anthropic says workloads that need to run in the US can use US-only inference at
1.1x pricing for input and output tokens. The headline prices on this page use standard global routing.Q · 07 What is Sonnet 5 best for? +
Anthropic positions Sonnet 5 for advanced coding, long-running agents, browser and computer use, enterprise workflows, customer-facing agents, and high-volume production tasks where Sonnet speed and cost matter.
Q · 08 Should I use Sonnet 5 instead of Sonnet 4.6? +
For new Claude API work, yes. Sonnet 5 is newer and permanently cheaper than Sonnet 4.6's
$3/$15 price band - the increase that would have levelled them on September 1 was cancelled. Keep Sonnet 4.6 only where you need an existing pinned migration target.