Anthropic: Claude Sonnet 5.5

Anthropic

flagship

Capabilities

VisionToolsStructured OutputExtended ThinkingLong Context

Strengths

Exceptional long-context handling with 1M token window
Strong vision and multimodal capabilities
Extended thinking support for complex reasoning
Structured output and tool use for enterprise workflows
High-quality code generation and bug fixing

Weaknesses

Extremely slow inference speed (0.07 score)
High output cost ($10/M tokens) limits high-volume applications
Closed-source weights and training data prevent self-hosting
Limited transparency on training methodology and data sources

Pricing

Input / 1M tokens

Output / 1M tokens

$10

Context window

1,000k

Transparency

Open weights

0.0 / 10

Open training data

0.0 / 10

Open methodology

3.5 / 10

Licence openness

1.0 / 10

Provider disclosure

5.5 / 10

FMTI company score

N/A

Composite:4.0 / 10

Closed-source model with limited public methodology disclosure. Anthropic publishes research on Constitutional AI and safety, but training data and detailed architectural decisions remain proprietary. Output pricing of $10/M tokens indicates flagship tier positioning.

Sustainability

Inference energy

N/A

Training footprint

N/A

Provider infrastructure

5.0 / 10

Composite:5.0 / 10

Anthropic's sustainability practices are not widely publicised. No confirmed renewable energy commitments or public efficiency metrics available. Conservative estimate pending provider disclosure.

Task Fitness

question answering

80%

code

84%

mathematics

50%

business communication

50%

vision

78%

analysis

80%

extraction

87%

long-form generation

82%

research

50%

reasoning

80%

summarisation

85%

translation

81%

conversation

83%