Anthropic: Claude Haiku 5.5

Not enough data yet to set these providers' speed against their price.

llm.buzz and Anthropic tie for the cheapest Anthropic: Claude Haiku 5.5, at $0.20 per million tokens blended.
llm.buzz serves Anthropic: Claude Haiku 5.5 fastest, at 193 tok/s (median over 24 hours).
llm.buzz starts answering Anthropic: Claude Haiku 5.5 quickest, first token in 2.48 s (median over 24 hours).
Anthropic has the best uptime for Anthropic: Claude Haiku 5.5, 100% over 72 hours.

ModelCompare with

Output speedtok/s, median over 24 h, higher is better

llm.buzz193 tok/s

AnthropicNot measured

First tokenseconds, median over 24 h, lower is better

llm.buzz2.48 s

Anthropic3.57 s

Blended price$ per 1M tokens, lower is better

llm.buzz$0.20

Anthropic$0.20

2 providers

Providers with their prices and measured figures. Select a column header to sort by it.

llm.buzz	$0.20	$0.10	$0.50	193 tok/stest	2.48 stest	98.6%test
Anthropic	$0.20	$0.10	$0.50	Not measured	3.57 stest	100%test

Customer traffic on this model is still below the publishing floor, so its figures are from llm.buzz's test requests, tagged "test".

The last 24 hours, by hour

Last 24 hours, from llm.buzz's test requests

Output speed, tok/s

Data table

Provider	From (UTC)	p50	p90	p99
llm.buzz	Oct 27, 01:00	193 tok/s	193 tok/s	193 tok/s
llm.buzz	Last 24 hours	193 tok/s	193 tok/s	193 tok/s

Time to first token

Data table

Provider	From (UTC)	p50	p90	p99
llm.buzz	Oct 27, 01:00	2.48 s	2.48 s	2.48 s
llm.buzz	Last 24 hours	2.48 s	2.48 s	2.48 s
Anthropic	Oct 26, 06:00	3.57 s	3.57 s	3.57 s
Anthropic	Oct 26, 11:00	3.57 s	3.57 s	3.57 s
Anthropic	Oct 26, 15:00	2.98 s	2.98 s	2.98 s
Anthropic	Oct 26, 20:00	3.57 s	3.57 s	3.57 s
Anthropic	Oct 27, 00:00	3.57 s	3.57 s	3.57 s
Anthropic	Last 24 hours	3.57 s	3.57 s	3.57 s

Last 72 hours, by hour

Best provider 100%

This model

llm.buzz

Anthropic

Speed and first tokenOutput speed and time to first token are medians over the last 24 hours, per provider, from text requests.
UptimeThe share of attempts the provider served over the 72 hours before the figures' end, provider failures counted against it.
Two sourcesCustomer traffic once enough organizations count toward a figure; below that, llm.buzz's test requests, short streamed requests sent to every live provider. The tables tag those "test".
PriceEach provider's price per million tokens, with a running discount applied. The blended price weighs input and output 3 to 1, a chat request's usual mix.

Let the router pickName the model; llm.buzz sends each request to a provider and moves to the next when one fails.