API pricing per million tokens
llama-3-1-8b-instruct Open weightsChecked 8h ago
- Input
- $0.02
- per 1M, DeepInfra (fp8)
- Output
- $0.04
- per 1M tokens
- Cached input
- n/a
- no batch tier listed
- Context
- 131K
- 118K max output
Where to buy it (5)
DeepInfrafp8, fp8$0.02 / $0.04$0.025 blended
Novita AIfp8, fp8$0.02 / $0.05$0.028 blended
Groqdefault region$0.05 / $0.08cached $0.025
Cloudflare Workers AIfp8, fp8$0.152 / $0.287$0.186 blended
CoreWeavebf16, bf16$0.22 / $0.22cached $0.22
Through a gateway
Your own numbers100M input and 20M output tokens a month: $2.80 direct.
Helicone$2.80+0%
Kong AI Gateway$2.80+0%
LiteLLM$2.80+0%
Vercel AI Gateway$2.80+0%
AWS Bedrock$2.80+0%
Azure AI Foundry$2.80+0%
Google Vertex AI$2.80+0%
Databricks Mosaic AI$2.80+0%
Hugging Face Inference$2.80+0%
Requesty$2.94+5%
Eden AI$2.95+5.5%
OpenRouter$2.95+5.5%
Snowflake Cortex AI$3.36+20%
Martian Gatewayn/a
Portkeyn/a
TrueFoundry AI Gatewayn/a
IBM watsonx.ain/a
OCI Generative AIn/a
NVIDIA NIMn/a
Related models
ModelInputOutputCachedContextReleased
Muse Spark 1.3Meta · 1M$1.25 / $4.25in / out per 1M$1.25$4.25$0.151MSep 2, 2026
Qwen3.7 FlashQwen (Alibaba) · 1M$0.03 / $0.13in / out per 1M$0.03$0.13$0.0061MJul 27, 2026
Text Embedding 3 SmallOpenAI · 8K$0.02 / Freein / out per 1M$0.02Freen/a8KOct 30, 2025
About Llama 3.1 8B Instruct
Meta's latest class of model (Llama 3.1) launched with a variety of sizes & flavors.
Takes text in and returns text. First listed Sep 23, 2026. Blended price $0.025 per 1M tokens at 3 input to 1 output.
Questions
How much does Llama 3.1 8B Instruct cost?
$0.02 per million input tokens and $0.04 per million output tokens.
What is the context window of Llama 3.1 8B Instruct?
131K tokens, with up to 118K tokens of output.
Where is Llama 3.1 8B Instruct cheapest?
Of 5 routes tracked, DeepInfra is cheapest at $0.02 input and $0.04 output per million tokens.
List prices from providers' pages on the date shown; your contract may differ. Logos via logo.dev; trademarks belong to their owners.