The AI arena is free today

Open Playground
API Provider9 active models1 organization

Anthropic: API pricing, speed & models

Anthropic hosts 9 active AI models, with input pricing from $1.00 per 1M tokens, with median throughput of 17 characters/sec, and P95 time to first token of 30.37s, with 88.1% success rate over 7 days. Compare Anthropic's API speed, pricing, and reliability against other inference providers.

First-party only
Median throughput17 c/s
Median TTFT3.47s
P95 TTFT30.37s
Success rate (7d)88.1%
From$1.00 /M tok

Catalog

Type
Price
22 models
Model
Claude Opus 5cache pricing
Claude Opus 5cache pricing
Claude Sonnet 5cache pricing · 42 QPS
Claude Sonnet 5cache pricing · 42 QPS
Claude Fable 5cache pricing
Claude Fable 5cache pricing
Claude Opus 4.8cache pricing · 42 QPS
Claude Opus 4.8cache pricing · 42 QPS
At a glance

Anthropicpricing, performance & catalog

The citable facts about Anthropic's 9 models — sourced from provider APIs and refreshed continuously.

Lowest price
Claude Haiku 4.5 at $1.00 per 1M input tokens
Highest median throughput
Claude Haiku 4.5 at 191 chars/s
Lowest median TTFT
Claude Sonnet 4.5 at 1.29s
Largest context
Claude Opus 5 at 1.0M tokens
Catalog
9 active models from 1 organization

FAQ

Common questions about Anthropic.

What is Anthropic?

Anthropic is an API provider that hosts large language models. Active models: 9; From (input): $1.00 / 1M tok; Median throughput: 17 c/s; Median TTFT: 3.47s; Success rate (7d): 88.1%.

How many models does Anthropic offer?

Anthropic currently serves 9 active models out of 20 historical offerings on LLM Stats.

What is Anthropic's API pricing?

Anthropic input pricing starts from $1.00 per 1M tokens, with the most expensive offering at $10 per 1M tokens. See the Pricing tab above for the full per-model breakdown.

How fast is Anthropic?

Anthropic delivers a median throughput of 17 characters per second and P95 time to first token of 30.37s across its model catalog. See the Models tab for per-model throughput and TTFT breakdowns.

Is Anthropic reliable?

Anthropic has a 88.1% success rate across 2.5K API calls in the last 7 days, with a 11.9% error rate.

Does Anthropic support function calling?

Yes. 22 of 9 models on Anthropic support function calling (tool use). The Capabilities tab lists which specific models accept tool definitions.

Does Anthropic support JSON mode and structured output?

Yes. 22 of 9 models on Anthropic support structured output (JSON mode / schema-constrained generation). The Capabilities tab shows which specific models accept response_format or json_schema parameters.

Does Anthropic offer batch inference?

Yes. 22 of 9 models on Anthropic support batch inference for cheaper, asynchronous workloads.

Does Anthropic support multimodal models?

Yes. Anthropic's catalog includes 9 vision-capable models. See the Models and Capabilities tabs for the full per-model breakdown.

Whose models does Anthropic host?

Anthropic hosts models from Anthropic. See the Models tab for the full catalog grouped by creator.

How do I start using Anthropic?

Sign up at https://anthropic.com to get an API key, then call Anthropic's API directly from your application. Most clients work out of the box by pointing the OpenAI SDK at Anthropic's base URL with your key. Use the Models and Pricing tabs above to pick the right model for your latency, cost, and context-window requirements.