Claude Sonnet 4.6
AnthropicAll Anthropic models
$3.00
$15.00
2026-02-17
Positioning: Balanced. Price level 3 of 4 among text models. Vendor-stated speed: Fast.
- Positioning
- Balanced, as the vendor describes it. The vendor has a newer generation.Source
- Price
- Level 3 of 4 among text models ($ is the cheapest), by output price per 1M tokens: $5 to under $20.
- Vendor-stated speed
- Fast, compared with other Claude models.Source
Three separate signals, not a quality score.
Sonnet all-rounder — computer-use automation and 1M-context coding sessions. Claude Sonnet 4.6 is an Anthropic text model on ATP Token, billed per 1M input and output tokens, with a lower cached-input rate.
ThinkingToolsVisionAudio
Per 1,000 short chats
$8.40
USD · 800/400 tokens
Input
$3.00
USD / 1M tokens
Output
$15.00
USD / 1M tokens
Released
Feb 17, 2026
Pricing
Pay for what you use: credits are deducted only when a request succeeds.
Input
$3.00
Cache read
$0.30
Output
$15.00
USD / 1M tokens
Input
$3.00
300 credits · USD / 1M tokens
Cache read
$0.30
30 credits · USD / 1M tokens
Output
$15.00
1,500 credits · USD / 1M tokens
Cache write · 5m TTL$3.75 / 1M
Cache write · 1h TTL$6.00 / 1M
Anthropic · Family · USD / 1M tokens
Text models from the same vendor on ATP. Bars fill up to the family's highest price.
| Model | Input | Output |
|---|---|---|
| Claude Fable 5claude-fable-5 | $10.00 | $50.00 |
| Claude Opus 4.7claude-opus-4-7 | $5.00 | $25.00 |
| Claude Opus 4.8claude-opus-4-8 | $5.00 | $25.00 |
| Claude Opus 5claude-opus-5 | $5.00 | $25.00 |
| Claude Sonnet 4.6 This pageclaude-sonnet-4-6 | $3.00 | $15.00 |
| Claude Sonnet 5claude-sonnet-5 | $3.00 | $15.00 |
| Claude Haiku 4.5claude-haiku-4-5 | $1.00 | $5.00 |
Cost per 1,000 typical requests
A quick estimate with fixed token counts per request. No caching, no retries. Try the estimator with your own numbers.
| Workload | USD / 1,000 |
|---|---|
| Short chat turn800 / 400 | $8.40 |
| RAG answer4,000 / 500 | $19.50 |
| Agent step with tools20,000 / 1,500 | $82.50 |
Benchmarks
Vendor-reported
- SWE-bench Verified79.6%
averaged over 10 trials (80.2% with a prompt modification)
Anthropic - Terminal-Bench 2.059.1%
Terminus-2 harness, thinking off; 1x guaranteed/3x ceiling resources, 5-15 samples per task
Anthropic - OSWorld-Verified72.5%
Agentic computer use (as reported at launch)
Anthropic - GPQA Diamond89.9%
Graduate-level reasoning
Anthropic
| Benchmark | Score | Source |
|---|---|---|
| SWE-bench Verifiedaveraged over 10 trials (80.2% with a prompt modification) | 79.6% | Anthropic |
| Terminal-Bench 2.0Terminus-2 harness, thinking off; 1x guaranteed/3x ceiling resources, 5-15 samples per task | 59.1% | Anthropic |
| OSWorld-VerifiedAgentic computer use (as reported at launch) | 72.5% | Anthropic |
| GPQA DiamondGraduate-level reasoning | 89.9% | Anthropic |
Scores as published by each vendor, not measured by ATP. Test conditions vary, so they are not directly comparable across vendors.
Same controls on every model: project keys, per-project model access, credit budgets, and a log of every request.Set up your organization
Cost estimator
Plug in your usage to see roughly what a month costs.
0200
0100
ESTIMATE
$60.00
USD / month
6,000 credits / month
- Input$30.00
- Output$30.00
$60.00 USD / month · 6,000 credits / month
Quickstart
Keep your SDK. Point it at ATP and use a project key. Read the docs
quickstart.py · OpenAI SDK
from openai import OpenAI
client = OpenAI(
base_url="https://api.atptoken.ai/v1",
api_key="atp-•••"
)
r = client.chat.completions.create(
model="claude-sonnet-4-6",
messages=[{"role":"user","content":"hi"}],
)Compare with similar models
Same type and positioning, from other vendors, closest in price first.
- GPT-5.4OpenAIPositioning: Balanced. Price level 3 of 4 among text models.$2.50 / $15.00In / out · per 1M tokens
- Kimi K2.6Moonshot (Kimi)Positioning: Balanced. Price level 2 of 4 among text models.$0.95 / $4.00In / out · per 1M tokens
- GLM-5Z.aiPositioning: Balanced. Price level 2 of 4 among text models.$1.00 / $3.20In / out · per 1M tokens
FAQ
- Can Claude Sonnet 4.6 use tools, read images, or think before answering?
- Claude Sonnet 4.6 supports tool calling, vision input, extended thinking and audio.
- Is cached input cheaper on Claude Sonnet 4.6?
- You pay per 1M input tokens and per 1M output tokens. Cached input costs 10% of the normal input price. Cache writes are priced separately. You only pay when a request succeeds. $1 = 100 credits. See pricing
- Is there a newer Anthropic model than Claude Sonnet 4.6 on ATP?
- Yes. Claude Opus 5 (claude-opus-5) came out on Jul 24, 2026, after Claude Sonnet 4.6 (Feb 17, 2026), and is the newer Anthropic text option on ATP. Both stay available: compare family prices on this page, then turn on the id you want for your project.
- How do I call Claude Sonnet 4.6 on ATP?
- Create a project API key and point any OpenAI-compatible SDK at https://api.atptoken.ai/v1. Set model to "claude-sonnet-4-6" (the API id, not the display name). Make sure the model is turned on for your project.
Get started
Start with $5. Credits never expire.
Create a project key, turn this model on, and pay only for what you use.