Anthropic Models

Q: Which Anthropic model should I start with?

claude-haiku-4-5 at $1.1/M input — the cheapest entry here that declares tool calling, and it carries a 200K context. Move up to claude-opus-4-1 when answer quality matters more than cost, or to claude-sonnet-5 for long-form reasoning.

Q: How is cached input billed?

The Cache read column is the rate for input tokens served from the prompt cache — for example claude-fable-5-1 bills cache hits at 2.5% of the input rate and anthropic-opus-4-6 bills cache hits at 10% of the input rate. Cache write is the surcharge for putting a prompt into the cache in the first place, and only a few upstreams bill it separately. A dash in either column means the catalog carries no cache rate for that model, so plan on paying the full input rate.

31 modelsGeneral models from $1.1/M inputUp to 1M context

Usage

Last 30 days · 2026-08-09 to 2026-09-07

Tokens

290B

Requests

Models in use

23 of 31

Tokens per day, stacked by model

claude-opus-5
claude-sonnet-5
claude-opus-4-8
claude-sonnet-4-6
claude-fable-5
claude-opus-4-7
claude-opus-4-6
claude-fable-5-1
11 more models

Which models that traffic went to

Share of 290B tokens. 4 models with traffic report no token counts and cannot be ranked here, including claude-3-7-sonnet and claude-3-sonnet-20240229 — they are in the request view.

The two views disagree on purpose: a model can take a large share of the calls and a small share of the tokens — many short requests — or the reverse. Which one matters depends on whether your cost is driven by call volume or by prompt length. Measured on AIHubMix over the last 30 days, counting the 31 model IDs listed on this page; traffic routed through upstream-specific IDs that are not in the public catalog is not included.

All 31 Anthropic Models

Open in model list

Anthropic models on AIHubMix with input and output modalities, context length, maximum output, price per million tokens including cache read and cache write rates, and measured throughput and latency.
	Modalities
claude-sonnet-5	Takes text, vision, returns text.	1M	128K	$2$10/M	$0.2/M	$2.5/M	7839 tok/s	1.03 s
claude-sonnet-4-6	Takes text, vision, returns text.	1M	64K	$3$15/M	$0.3/M	$3.75/M	69 tok/s	0.88 s
claude-sonnet-4-6-think	Takes text, vision, returns text.	1M	32K	$3$15/M	$0.3/M	$3.75/M	69 tok/s	1.18 s
claude-sonnet-4-5	Takes text, vision, returns text.	1M	64K	$3.3$16.5/M	$0.33/M	$4.125/M	43 tok/s	1.62 s
claude-sonnet-4-5-think	Takes text, vision, returns text.	1M	64K	$3.3$16.5/M	$0.33/M	$4.125/M	—	—
claude-opus-4-6	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	60 tok/s	1.21 s
claude-opus-4-6-think	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	51 tok/s	1.74 s
claude-opus-4-7	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	81 tok/s	1.22 s
claude-opus-4-7-think	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	90 tok/s	1.17 s
claude-opus-4-8	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	83 tok/s	1.05 s
claude-opus-4-8-think	Takes text, vision, returns text.	1M	32K	$5$25/M	$0.5/M	$6.25/M	54 tok/s	4.61 s
claude-opus-5	Takes text, vision, returns text.	1M	128K	$5$25/M	$0.5/M	$6.25/M	62 tok/s	1.42 s
claude-fable-5	Takes text, vision, returns text.	1M	128K	$11$55/M	$1.1/M	$13.75/M	63 tok/s	2.41 s
claude-fable-5-1	Takes text, vision, returns text.	1M	128K	$11$55/M	$0.275/M	$13.75/M	51 tok/s	8.09 s
claude-3-haiku-20240229	Takes text, vision, returns text.	200K	—	$0.275$0.275/M	—	—	67 tok/s	0.80 s
claude-3-haiku@20240307	Takes text, vision, returns text.	200K	—	$0.275$1.375/M	—	—	67 tok/s	0.80 s
claude-haiku-4-5	Takes text, vision, returns text.	200K	131K	$1.1$5.5/M	$0.11/M	$1.375/M	153 tok/s	0.51 s
claude-3-7-sonnet	Takes text, vision, returns text.	200K	128K	$3.3$16.5/M	—	—	57 tok/s	0.63 s
claude-3-sonnet-20240229	Takes text, vision, returns text.	200K	—	$3.3$16.5/M	—	—	67 tok/s	0.76 s
anthropic-opus-4-6	Takes text, vision, returns text.	200K	32K	$5$25/M	$0.5/M	$6.25/M	40 tok/s	3.07 s
claude-opus-4-5	Takes text, vision, returns text.	200K	32K	$5$25/M	$0.5/M	$6.25/M	49 tok/s	1.52 s
claude-opus-4-5-think	Takes text, vision, returns text.	200K	32K	$5$25/M	$0.5/M	$6.25/M	—	—
claude-opus-4-1	Takes text, vision, returns text.	200K	32K	$16.5$82.5/M	—	—	41 tok/s	1.69 s
claude-3-haiku-20240307	Takes text, vision, returns text.	—	—	$0.275$1.375/M	—	—	67 tok/s	0.80 s
claude-instant-1.2		—	—	$0.88$3.96/M	—	—	—	—
claude-instant-1		—	—	$1.793$1.793/M	—	—	—	—
claude-2		—	—	$8.8$8.8/M	—	—	—	—
claude-2.0		—	—	$8.8$39.6/M	—	—	—	—
claude-2.1		—	—	$8.8$39.6/M	—	—	—	—
claude-3-opus-20240229		—	—	$16.5$82.5/M	—	—	—	—
claude-3-opus@20240229		—	—	$16.5$82.5/M	—	—	—	—

Prices are USD per million tokens; cache read and cache write are the rates for prompt-cache hits and for writing a prompt into the cache. Throughput and latency are measured on AIHubMix — the same figures the model detail page shows — not vendor claims. A dash means the catalog does not publish that field for that model, which is not the same as the model not supporting it.

Common Comparisons

anthropic-opus-4-6 vs claude-3-7-sonnetanthropic-opus-4-6 vs claude-3-haiku-20240229anthropic-opus-4-6 vs claude-3-haiku-20240307anthropic-opus-4-6 vs qwen3.8-max-previewanthropic-opus-4-6 vs kimi-k3

Anthropic on AIHubMix

Which Anthropic model should I start with?

claude-haiku-4-5 at $1.1/M input — the cheapest entry here that declares tool calling, and it carries a 200K context. Move up to claude-opus-4-1 when answer quality matters more than cost, or to claude-sonnet-5 for long-form reasoning.

Which of these models reason before answering?

20 of the 31 models here declare a reasoning phase — they work through the problem before producing an answer, which helps on multi-step problems at the cost of extra output tokens. Use the Reasoning filter above the table to see them. The catalog does not record anything further about how they differ, so this page does not sort them into families.

Why are there several entries for the same model?

Because each row is a route you can call, not a model release. Some IDs name an upstream (azure-, alicloud-, cc-), and some differ only in capitalisation, kept so older integrations keep working.

The catalog does not carry a field saying which of those a given row is, so this page does not sort them into buckets it would have to invent. Every row shows that route’s own price, context and speed — compare those directly, and open a model to see the upstreams that serve it.

How is cached input billed?

The Cache read column is the rate for input tokens served from the prompt cache — for example claude-fable-5-1 bills cache hits at 2.5% of the input rate and anthropic-opus-4-6 bills cache hits at 10% of the input rate. Cache write is the surcharge for putting a prompt into the cache in the first place, and only a few upstreams bill it separately. A dash in either column means the catalog carries no cache rate for that model, so plan on paying the full input rate.

Do I need a separate Anthropic account?

No. One AIHubMix key covers every model on this page, and switching between them is a change to the model string — billing, rate limits, and logs stay in one place.