4032
provider brief

Anthropic

← Back to atlas
2 models · 3 modalities · 2 tiers

market

Anthropic

Anthropic lineup overview: capabilities, latency profiles, and where each model fits inside the 4032.ai bridge.

Modalities

Code · Text · Vision

Coverage across the lineup.

Max context

200k tokens

Largest window offered by this provider.

Tiers

flagship · fast

Blend of speed, reasoning, and openness.

lineup

Anthropic models

Compare the models from Anthropic side by side. Look at tiers, latency, pricing, and where they slot into your workloads.

2024 flagship Balanced; responsive for chat and workflow calls

Claude 3.7 Sonnet

Anthropic's fast flagship model with strong writing quality, safety alignment, and broad domain coverage.

Details →

Context

200k tokens

Modalities

Text · Vision · Code

Pricing

$3.00 / 1M input tokens, $15.00 / 1M output tokens

Availability

Anthropic API, Amazon Bedrock, Google Cloud Vertex

Strengths

  • High-quality narrative writing and summarization with an even tone.
  • Grounded responses with reliable refusals and safety guardrails.
  • Consistent tool-use behavior for multi-step agent plans.

Best for

  • Enterprise copilots that need predictable tone and safety posture.
  • Long-form rewriting, knowledge distillation, and doc QA.
  • Ops workflows that mix structured output with human approvals.
2024 fast Low; tuned for interactive experiences

Claude 3.5 Haiku

Lightweight Claude tuned for speed while keeping surprisingly strong reasoning and extraction quality.

Details →

Context

200k tokens

Modalities

Text · Vision · Code

Pricing

$0.80 / 1M input tokens, $4.00 / 1M output tokens

Availability

Anthropic API, Amazon Bedrock, Google Cloud Vertex

Strengths

  • Fast responses with grounded summaries and low hallucination rates.
  • Excellent structured extraction across tables, docs, and receipts.
  • Cost-efficient while still supporting reliable tool use.

Best for

  • High-traffic chat surfaces and internal assistants.
  • Knowledge extraction, enrichment, and tagging.
  • Routing and pre-processing before heavier models or pipelines.