OpenAI: GPT-6 Astra

OpenAI and llm.buzz tie for the cheapest, and OpenAI is the fastest provider of this model. llm.buzz costs the same as a provider on the line, for less speed.

All benchmarks Model page Try in Playground

Live providers: 2; Same model, one API
Cheapest: $20.00 blended; llm.buzz
Fastest output: 31 tok/s; OpenAI
Quickest first token: 1.72 s; OpenAI
Best uptime: 100%; OpenAI

llm.buzz and OpenAI tie for the cheapest OpenAI: GPT-6 Astra, at $20.00 per million tokens blended.
OpenAI serves OpenAI: GPT-6 Astra fastest, at 31 tok/s (median over 24 hours).
OpenAI starts answering OpenAI: GPT-6 Astra quickest, first token in 1.72 s (median over 24 hours).
OpenAI has the best uptime for OpenAI: GPT-6 Astra, 100% over 72 hours.

ModelCompare with

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

OpenAI31 tok/s

llm.buzz18 tok/s

First tokenseconds, median over 24 h, lower is better

OpenAI1.72 s

llm.buzz2.48 s

Blended price$ per 1M tokens, lower is better

llm.buzz$20.00

OpenAI$20.00

2 providers

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

llm.buzz	$20.00	$10.00	$50.00	18 tok/stest	2.48 stest	97.1%test
OpenAI	$20.00	$10.00	$50.00	31 tok/stest	1.72 stest	100%test

Speed against priceOutput speed in tok/s (median over 24 h), against the blended pricenobody beats these on both price and speed

0153045$10.00$15.00$20.00$25.00$30.00

llm.buzz$20.00 blended, 18 tok/s

OpenAI$20.00 blended, 31 tok/s

Blended price, $ per million tokens (3 input to 1 output)

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

llm.buzz18 tok/s
OpenAI31 tok/s

Data table

Provider	From (UTC)	p50	p90	p99
llm.buzz	Oct 26, 08:00	18 tok/s	18 tok/s	18 tok/s
llm.buzz	Oct 26, 12:00	93 tok/s	93 tok/s	93 tok/s
llm.buzz	Oct 26, 16:00	15 tok/s	15 tok/s	15 tok/s
llm.buzz	Oct 26, 20:00	54 tok/s	54 tok/s	54 tok/s
llm.buzz	Oct 27, 00:00	18 tok/s	18 tok/s	18 tok/s
llm.buzz	Oct 27, 04:00	15 tok/s	15 tok/s	15 tok/s
llm.buzz	Last 24 hours	18 tok/s	15 tok/s	15 tok/s
OpenAI	Oct 26, 06:00	31 tok/s	31 tok/s	31 tok/s
OpenAI	Oct 26, 10:00	31 tok/s	31 tok/s	31 tok/s
OpenAI	Oct 26, 15:00	54 tok/s	54 tok/s	54 tok/s
OpenAI	Oct 26, 19:00	37 tok/s	37 tok/s	37 tok/s
OpenAI	Oct 26, 23:00	22 tok/s	22 tok/s	22 tok/s
OpenAI	Oct 27, 03:00	31 tok/s	31 tok/s	31 tok/s
OpenAI	Last 24 hours	31 tok/s	22 tok/s	22 tok/s

Time to first token

llm.buzz2.48 s
OpenAI1.72 s

Data table

Provider	From (UTC)	p50	p90	p99
llm.buzz	Oct 26, 08:00	3.57 s	3.57 s	3.57 s
llm.buzz	Oct 26, 12:00	2.48 s	2.48 s	2.48 s
llm.buzz	Oct 26, 16:00	2.48 s	2.48 s	2.48 s
llm.buzz	Oct 26, 20:00	2.98 s	2.98 s	2.98 s
llm.buzz	Oct 27, 00:00	2.98 s	2.98 s	2.98 s
llm.buzz	Oct 27, 04:00	2.48 s	2.48 s	2.48 s
llm.buzz	Last 24 hours	2.48 s	3.57 s	3.57 s
OpenAI	Oct 26, 06:00	1.20 s	1.20 s	1.20 s
OpenAI	Oct 26, 10:00	2.98 s	2.98 s	2.98 s
OpenAI	Oct 26, 15:00	5.14 s	5.14 s	5.14 s
OpenAI	Oct 26, 19:00	3.57 s	3.57 s	3.57 s
OpenAI	Oct 26, 23:00	1.72 s	1.72 s	1.72 s
OpenAI	Oct 27, 03:00	1.44 s	1.44 s	1.44 s
OpenAI	Last 24 hours	1.72 s	5.14 s	5.14 s

Uptime

Last 72 hours, by hour

Best provider 100%

This model

llm.buzz

OpenAI

Up: 95% or more served
Degraded: 85% to 95%
Down: below 85%
No data
From llm.buzz's test requests

How we measure

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.

Get API key