NOUMENOpen app
NOUMEN

Anthropic · Frontier

Claude 4.7 Opus Thinking

Added Apr 16, 2026

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Use in the app

Claude Opus 4.7 with thinking enabled (default budget: 16k tokens).

  • 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
$5.00
Output
$25.00
Cache read
$0.50

Reference prices as published by the model provider. API example

Specifications

Context window
1M
Max output
128K
Average output (7 days)
1.9K tokens
Longer than 90% of 473 models

Side by side

Claude 4.7 Opus Thinking compared

Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.

Compare models

Choose up to 3 models to compare.1 of 3 chosen

Loading models…

Claude 4.7 Opus ThinkingAnthropicThis modelGemini 3.8 FlashGoogleRemove
ProviderAnthropicGoogle
Discovery tierFrontierEfficient
AddedApr 16, 2026Sep 2, 2026
Limits
Context window1M1.05MBest
Max output128KBest65.54K
Average output1.9K tokens1.2K tokens
Pricing, USD per 1M tokens
Input$5.00$0.75Best
Output$25.00$3.75Best
Cache read$0.50$0.075Best
Cache write—$0.0417
Performance
Throughput98.9 tok/s362.4 tok/sBest
Time to first token3.0 sBest9.2 s
Request success (24h)99.1%99.8%Best
Benchmark indexes
Intelligence Index40.7Best39.8
Coding Index73.674.1Best
Capabilities
ReasoningYesYes
ToolsYesYes
VisionYesYes
Structured outputYesYes
PDF inputYesYes
Audio inputNoYes
Video inputNoYes
Reasoning effortlow, medium, high, xhigh, maxminimal, low, medium, high

Public evaluations

Benchmarks

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

  • Claude 4.7 Opus Thinking
  • Gemini 3.8 Flash

Intelligence Index

  • Claude 4.7 Opus Thinking40.7
  • Gemini 3.8 Flash39.8

Claude 4.7 Opus Thinking tested at max reasoning effort

Claude 4.7 Opus Thinking: Better than 85% of 326 scored models

3.8Catalog models by score55.1

Indexes

  • Coding Index

    Composite of the coding evaluations

    • Claude 4.7 Opus Thinking: 73.6
    • Gemini 3.8 Flash: 74.1Top

Agentic work

  • AutomationBench-AA

    Workflow automation, penalized for guardrail violations

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 60.9%

  • AutomationBench-AA tasks completed

    Workflows finished end to end without a guardrail violation

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 31.1%

  • AA-Briefcasev1.1

    Agentic knowledge work, rated head to head

    • Claude 4.7 Opus Thinking: 1,252 EloTop
    • Gemini 3.8 Flash: 1,181 Elo

  • GDPval-AAv2.1

    Economically valuable occupational tasks, rated head to head

    • Claude 4.7 Opus Thinking: 1,353 Elo
    • Gemini 3.8 Flash: 1,423 EloTop

Document reasoning

  • GDP.pdf

    Professional reasoning over PDF documents, all-pass rate

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 22.8%

  • AA-LCRv1.1

    Reasoning across long documents

    • Claude 4.7 Opus Thinking: 78.7%
    • Gemini 3.8 Flash: 84.0%Top

  • MLCR-AA

    Long-context reasoning over medical records

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 12.8%

Reasoning and knowledge

  • Humanity’s Last Exam

    Expert-level questions across academic fields

    • Claude 4.7 Opus Thinking: 42.3%Top
    • Gemini 3.8 Flash: 42.1%

  • GPQA Diamond

    Graduate-level science questions

    • Claude 4.7 Opus Thinking: 91.4%
    • Gemini 3.8 Flash: 93.5%Top

  • IFBench

    Following precise output instructions

    • Claude 4.7 Opus Thinking: 58.6%
    • Gemini 3.8 Flash: —

Coding

  • Terminal-Benchv4.0

    Practical software engineering in a terminal

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 19.7%

  • SciCode

    Scientific computing in Python

    • Claude 4.7 Opus Thinking: —
    • Gemini 3.8 Flash: 55.1%

Legacy evaluations

  • Terminal-Bench Hard

    Earlier terminal task suite

    • Claude 4.7 Opus Thinking: 51.5%
    • Gemini 3.8 Flash: —

  • τ²-Bench Telecom

    Tool use in a customer-support simulation

    • Claude 4.7 Opus Thinking: 88.6%
    • Gemini 3.8 Flash: —

  • AA-LCR (unversioned)

    Long-context reasoning, earlier grading

    • Claude 4.7 Opus Thinking: 78.7%
    • Gemini 3.8 Flash: 84.0%Top

Scores from Artificial Analysis, updated Oct 26, 2026. Claude 4.7 Opus Thinking was tested at max 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
98.9 tok/s
Faster than 62% of 256 models
Time to first token
3.0 s
Quicker than 51% of 256 models
Prompt cache hits
40.2%
About $3.19 per 1M input tokens with caching
Request success (24h)
98.7%
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 85.7%
  • 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-opus-4.7:thinking",
    "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-opus-4.7:thinking",
    messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)

Anthropic

Other Anthropic models, starting with the same family.