Name: Meta: Llama 3.1 8B Instruct API
Brand: Meta Llama
SKU: meta-llama/llama-3.1-8b-instruct

Best price — Output ($/1M tokens)

Markets

One row per access path (venue → underlying provider). Prices are each venue's published list price; performance is measured by Artificial Analysis. Your actual bill may vary by routing settings, fallbacks, region / ZDR, long-context tiers, and request features (JSON mode, tools, vision).

#	Provider	Class	Input $/M	Output $/M	Cache ($/M Read)	vs consensus	TTFT	Throughput	Uptime 24h	Requests 24h
1	google-vertex	Direct host	free	free	—	-100.0% vs consensus	—	—	—	—
2	openrouter → deepinfra	Reseller	$0.02	$0.04	—	-60.0% vs consensus	—	—	—	—
3	deepinfra → meta	Reseller	$0.02	$0.05	—	-50.0% vs consensus	—	—	—	—
4	novita → meta	Reseller	$0.02	$0.05	—	-50.0% vs consensus	—	—	—	—
5	openrouter → groq	Reseller	$0.05	$0.08	$0.03	-20.0% vs consensus	—	—	—	—
6	openrouter → meta	Reseller	$0.05	$0.08	$0.03	-20.0% vs consensus	—	—	—	—
7	ovhcloud	Direct host	$0.1	$0.1	—	+0.0% vs consensus	—	—	—	—
8	amazon-bedrock	Direct host	$0.22	$0.22	—	+120.0% vs consensus	—	—	—	—
9	bedrock	Direct host	$0.22	$0.22	—	+120.0% vs consensus	—	—	—	—
10	openrouter → coreweave	Reseller	$0.22	$0.22	$0.22	+120.0% vs consensus	—	—	—	—
11	wandb → meta	Reseller	$0.22	$0.22	—	+120.0% vs consensus	—	—	—	—
12	openrouter → cloudflare	Reseller	$0.15	$0.29	—	+187.0% vs consensus	—	—	—	—
13	oci	Direct host	$0.72	$0.72	—	+620.0% vs consensus	—	—	—	—
14	cloudflare Disabled by cloudflare · stale	Direct host	free	free	—	-100.0% vs consensus	702ms	202.2 tps	—	—
15	deepinfra Disabled by deepinfra · disabled · bfloat16	Direct host	$0.02	$0.05	—	-50.0% vs consensus	940ms	35.4 tps	—	—
16	novita Disabled by novita · stale	Direct host	$0.02	$0.05	—	-50.0% vs consensus	917ms	162.4 tps	—	—
17	openrouter → novita Disabled by openrouter · disabled · fp8	Reseller	$0.02	$0.05	—	-50.0% vs consensus	—	—	—	—
18	groq Disabled by groq · stale	Direct host	$0.05	$0.08	—	-20.0% vs consensus	971ms	642.0 tps	—	—
19	compactifai Disabled by compactifai · stale	Direct host	$0.02	$0.09	—	-10.0% vs consensus	1.2s	153.1 tps	—	—
20	friendli-ai Disabled by friendli-ai · stale	Direct host	$0.1	$0.1	—	+0.0% vs consensus	726ms	247.4 tps	—	—
21	sambanova Disabled by sambanova · stale	Direct host	$0.1	$0.2	—	+100.0% vs consensus	—	—	—	—
22	coreweave Disabled by coreweave · stale	Direct host	$0.22	$0.22	—	+120.0% vs consensus	647ms	141.1 tps	—	—
23	openrouter → wandb Disabled by openrouter · disabled · bf16	Reseller	$0.22	$0.22	$0.22	+120.0% vs consensus	—	—	—	—
24	openrouter → wandb-legacy Disabled by openrouter · disabled · bf16	Reseller	$0.22	$0.22	$0.22	+120.0% vs consensus	—	—	—	—
25	scaleway Disabled by scaleway · stale	Direct host	$0.23	$0.23	—	+130.0% vs consensus	968ms	141.4 tps	—	—
26	databricks Disabled by databricks · stale	Direct host	$0.15	$0.45	—	+350.0% vs consensus	866ms	151.9 tps	—	—
27	azure → azure-ai Disabled by azure · stale	Reseller	$0.3	$0.61	—	+510.0% vs consensus	—	—	—	—

#
1
Provider
google-vertex
Class
Direct host
$ / 1M tokens
Input $ / 1M tokens
free
Output $ / 1M tokens
free
Cache read $ / 1M tokens
—
-100.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
2
Provider
openrouter → deepinfra
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.04
Cache read $ / 1M tokens
—
-60.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
3
Provider
deepinfra → meta
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.05
Cache read $ / 1M tokens
—
-50.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
4
Provider
novita → meta
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.05
Cache read $ / 1M tokens
—
-50.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
5
Provider
openrouter → groq
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.05
Output $ / 1M tokens
$0.08
Cache read $ / 1M tokens
$0.03
-20.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
6
Provider
openrouter → meta
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.05
Output $ / 1M tokens
$0.08
Cache read $ / 1M tokens
$0.03
-20.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
7
Provider
ovhcloud
Class
Direct host
$ / 1M tokens
Input $ / 1M tokens
$0.1
Output $ / 1M tokens
$0.1
Cache read $ / 1M tokens
—
+0.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
8
Provider
amazon-bedrock
Class
Direct host
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
—
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
9
Provider
bedrock
Class
Direct host
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
—
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
10
Provider
openrouter → coreweave
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
$0.22
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
11
Provider
wandb → meta
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
—
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
12
Provider
openrouter → cloudflare
Class
Reseller
$ / 1M tokens
Input $ / 1M tokens
$0.15
Output $ / 1M tokens
$0.29
Cache read $ / 1M tokens
—
+187.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
13
Provider
oci
Class
Direct host
$ / 1M tokens
Input $ / 1M tokens
$0.72
Output $ / 1M tokens
$0.72
Cache read $ / 1M tokens
—
+620.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
14
Provider
cloudflare
Class
Direct host
Disabled by cloudflare · stale
$ / 1M tokens
Input $ / 1M tokens
free
Output $ / 1M tokens
free
Cache read $ / 1M tokens
—
-100.0% vs consensus
TTFT
702ms
Throughput
202.2 tps
Uptime 24h
—
Requests 24h
—
#
15
Provider
deepinfra
Class
Direct host
Disabled by deepinfra · disabled · bfloat16
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.05
Cache read $ / 1M tokens
—
-50.0% vs consensus
TTFT
940ms
Throughput
35.4 tps
Uptime 24h
—
Requests 24h
—
#
16
Provider
novita
Class
Direct host
Disabled by novita · stale
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.05
Cache read $ / 1M tokens
—
-50.0% vs consensus
TTFT
917ms
Throughput
162.4 tps
Uptime 24h
—
Requests 24h
—
#
17
Provider
openrouter → novita
Class
Reseller
Disabled by openrouter · disabled · fp8
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.05
Cache read $ / 1M tokens
—
-50.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
18
Provider
groq
Class
Direct host
Disabled by groq · stale
$ / 1M tokens
Input $ / 1M tokens
$0.05
Output $ / 1M tokens
$0.08
Cache read $ / 1M tokens
—
-20.0% vs consensus
TTFT
971ms
Throughput
642.0 tps
Uptime 24h
—
Requests 24h
—
#
19
Provider
compactifai
Class
Direct host
Disabled by compactifai · stale
$ / 1M tokens
Input $ / 1M tokens
$0.02
Output $ / 1M tokens
$0.09
Cache read $ / 1M tokens
—
-10.0% vs consensus
TTFT
1.2s
Throughput
153.1 tps
Uptime 24h
—
Requests 24h
—
#
20
Provider
friendli-ai
Class
Direct host
Disabled by friendli-ai · stale
$ / 1M tokens
Input $ / 1M tokens
$0.1
Output $ / 1M tokens
$0.1
Cache read $ / 1M tokens
—
+0.0% vs consensus
TTFT
726ms
Throughput
247.4 tps
Uptime 24h
—
Requests 24h
—
#
21
Provider
sambanova
Class
Direct host
Disabled by sambanova · stale
$ / 1M tokens
Input $ / 1M tokens
$0.1
Output $ / 1M tokens
$0.2
Cache read $ / 1M tokens
—
+100.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
22
Provider
coreweave
Class
Direct host
Disabled by coreweave · stale
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
—
+120.0% vs consensus
TTFT
647ms
Throughput
141.1 tps
Uptime 24h
—
Requests 24h
—
#
23
Provider
openrouter → wandb
Class
Reseller
Disabled by openrouter · disabled · bf16
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
$0.22
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
24
Provider
openrouter → wandb-legacy
Class
Reseller
Disabled by openrouter · disabled · bf16
$ / 1M tokens
Input $ / 1M tokens
$0.22
Output $ / 1M tokens
$0.22
Cache read $ / 1M tokens
$0.22
+120.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—
#
25
Provider
scaleway
Class
Direct host
Disabled by scaleway · stale
$ / 1M tokens
Input $ / 1M tokens
$0.23
Output $ / 1M tokens
$0.23
Cache read $ / 1M tokens
—
+130.0% vs consensus
TTFT
968ms
Throughput
141.4 tps
Uptime 24h
—
Requests 24h
—
#
26
Provider
databricks
Class
Direct host
Disabled by databricks · stale
$ / 1M tokens
Input $ / 1M tokens
$0.15
Output $ / 1M tokens
$0.45
Cache read $ / 1M tokens
—
+350.0% vs consensus
TTFT
866ms
Throughput
151.9 tps
Uptime 24h
—
Requests 24h
—
#
27
Provider
azure → azure-ai
Class
Reseller
Disabled by azure · stale
$ / 1M tokens
Input $ / 1M tokens
$0.3
Output $ / 1M tokens
$0.61
Cache read $ / 1M tokens
—
+510.0% vs consensus
TTFT
—
Throughput
—
Uptime 24h
—
Requests 24h
—

Source ecosystem

Loading…