Claude Sonnet 4.5
Released September 2025 and optimized for agents and coding, approaching Opus-level performance within the Sonnet tier.
Key Specifications
- Context window
- 200K tokens
- Max output
- 64K tokens
- Knowledge cutoff
- web search available
- Released
- Inputs
- Text, image
- Capabilities Show details
- Reasoning Thinks before it answers, always on or as a switchable mode. Function calling Connect to external tools, APIs, and systems. Structured output Return responses in structured formats like JSON. Web search Search the web for up-to-date information. Code execution Write and execute code in a sandboxed environment. File upload Process uploaded documents, spreadsheets, and other files.
Hosted API pricing
| Provider | Input /1M tokens | Output /1M tokens | Cached input /1M | Cost at 10M in + 2M out | |
|---|---|---|---|---|---|
|
|
$3.00 | $15.00 | $0.30 | $60.00 | View |
|
|
$3.00 | $15.00 | $0.30 | $60.00 | View |
|
|
$3.00 | $15.00 | $60.00 | View |
Heads up: Base-tier, on-demand rates per 1M tokens; cached, batch and long-context tiers excluded. A provider may serve a shorter context or a quantized build than the creator's release. Verify before provisioning. More on how we price.
Similarly priced models
The models nearest Claude Sonnet 4.5 by blended rate, each at its own cheapest provider.
| Model | Blended /1M | Input /1M | Output /1M | Context | Cutoff | vs Claude Sonnet 4.5 |
|---|---|---|---|---|---|---|
|
|
$3.75 | $2.50 | $10.00 | 128K | Feb 2023 | −25% |
|
|
$3.75 | $2.50 | $10.00 | 128K | Oct 2023 | −25% |
|
|
$3.7917 | $1.75 | $14.00 | 400K | Aug 2025 | −24% |
|
|
$3.7917 | $1.75 | $14.00 | 400K | Aug 2025 | −24% |
|
|
$4.5833 | $2.50 | $15.00 | 1M | Aug 2025 | −8% |
|
|
$5.00 | $3.00 | $15.00 | 200K | Jan 2025 | |
|
|
$5.00 | $3.00 | $15.00 | 1M | Aug 2025 | 0% |
|
|
$8.3333 | $5.00 | $25.00 | 200K | May 2025 | +67% |
|
|
$8.3333 | $5.00 | $25.00 | 1M | May 2025 | +67% |
|
|
$8.3333 | $5.00 | $25.00 | 1M | Jan 2026 | +67% |
|
|
$8.3333 | $5.00 | $25.00 | 1M | Jan 2026 | +67% |
Prices are USD per 1M tokens at each model's cheapest listed provider. Blended is the cost of 10M input plus 2M output tokens, spread over the 12M.
Frequently Asked Questions
What is Claude Sonnet 4.5 good for?
Agents and coding, taking text, images and PDFs. It is priced at Anthropic's Sonnet tier, below Opus.
When is Claude Sonnet 4.5 not a good fit?
Very long inputs and long single answers, where Sonnet 4.6 takes much more of both at the same price and Sonnet 5 at a lower one. No audio or video input, and closed weights, so no self-hosting.
What is the cheapest way to run Claude Sonnet 4.5?
The cheapest hosted rate is $3.00 / $15.00 per 1M tokens via Anthropic.
How does Claude Sonnet 4.5 compare with Claude Sonnet 4.6?
Anthropic calls Claude Sonnet 4.6 a full upgrade over Claude Sonnet 4.5 across coding, computer use, long-context reasoning, agent planning and knowledge work, with better consistency and instruction following. It says Sonnet 4.6 also resists prompt injection much better. Same price and text and image input. Claude Sonnet 4.5 came out 5 months earlier and has an older knowledge cutoff (January 2025 against August 2025), a smaller context (200K against 1M tokens) and a smaller max output (64K against 128K tokens).
Can I self-host Claude Sonnet 4.5?
No. Anthropic does not publish the weights; Claude Sonnet 4.5 is available only through hosted APIs.
More from Anthropic
| Model | Context | Input /1M | Output /1M |
|---|---|---|---|
|
|
1M | $3.00 | $15.00 |
|
|
200K | $3.00 | $15.00 |
|
|
1M | $2.00 | $10.00 |
|
|
1M | $5.00 | $25.00 |
|
|
1M | $5.00 | $25.00 |
|
|
1M | $5.00 | $25.00 |
|
|
1M | $5.00 | $25.00 |
|
|
200K | $5.00 | $25.00 |