Anthropic · Frontier

Claude Sonnet 4.6

Added Feb 17, 2026

Claude Sonnet 4.6 is Anthropic's most capable Sonnet yet — a full upgrade across coding, computer use, long-context reasoning, agent planning, and design. Supports up to a 1M-token context window at Sonnet pricing.

Reasoning
Tools
Vision
Structured output
PDF input

Input: Text, Images
Output: Text
Reasoning effort: low · medium · high · xhigh · max
Discovery tier: Frontier

Pricing

USD per 1M tokens

Input: $3.00
Output: $15.00
Cache read: $0.30

Reference prices as published by the model provider. API example

Specifications

Context window: 1M
Max output: 128K
Average output (7 days): 401 tokens; Longer than 40% of 473 models

Side by side

Compare models

Put Claude Sonnet 4.6 next to up to three other models. The comparison stays in the link, so you can share it.

TryDeepSeek V4.1 Flash 1017 Ling 3.0 Flash VL Inkling All 3

Public evaluations

Benchmarks

Independent evaluation results, with how each score ranks among the models in the Noumen catalog.

Intelligence Index

24.7

Tested at high reasoning effort

Better than 60% of 326 scored models

3.8Catalog models by score55.1

Document reasoning

AA-LCRv1.1
68.3%
Reasoning across long documents
Better than 44% of 309 scored models

Reasoning and knowledge

Humanity’s Last Exam
13.3%
Expert-level questions across academic fields
Better than 39% of 323 scored models
GPQA Diamond
79.9%
Graduate-level science questions
Better than 48% of 291 scored models
IFBench
41.2%
Following precise output instructions
Better than 29% of 215 scored models

Legacy evaluations

Terminal-Bench Hard
46.2%
Earlier terminal task suite
Better than 88% of 212 scored models
τ²-Bench Telecom
79.5%
Tool use in a customer-support simulation
Better than 60% of 214 scored models
AA-LCR (unversioned)
68.3%
Long-context reasoning, earlier grading
Better than 44% of 309 scored models

Scores from Artificial Analysis, updated Oct 26, 2026. Claude Sonnet 4.6 was tested at high reasoning effort; compare scores measured at the same effort.Artificial Analysis

Live routing data

Performance and reliability

Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.

Throughput: 52.4 tok/s; Faster than 35% of 260 models
Time to first token: 1.7 s; Quicker than 84% of 260 models
Prompt cache hits: 37.1%; About $2.00 per 1M input tokens with caching
Request success (24h): 99.1%; Reliable

Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.

Hourly request success · last 72 hoursLowest hour 65.8%

Healthy (98% or more)
Degraded (90–98%)
Down (under 90%)
No requests

Developers

Use it through the API

Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.

Read the quickstart

curl https://www.noumen.xyz/api/v1/chat/completions \
  -H "Authorization: Bearer $NOUMEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "anthropic/claude-sonnet-4.6",
    "reasoning_effort": "medium",
    "messages": [{ "role": "user", "content": "Explain this stack trace." }]
  }'

import os
from openai import OpenAI

client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
    model="anthropic/claude-sonnet-4.6",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Anthropic

Other Anthropic models, starting with the same family.

Document reasoning

AA-LCRv1.1

Reasoning and knowledge

Humanity’s Last Exam

GPQA Diamond

IFBench

Legacy evaluations

Terminal-Bench Hard

τ²-Bench Telecom

AA-LCR (unversioned)