Skip to main content

Groq

Inference host

Groq runs open and licensed models on its own hardware and sells them by usage, often faster or cheaper than the maker.

PricingDocs
Visit groq.com
Models10
Cheapest for
0 models
Serves
3 model types
Visits in September 20263.6M−13.6%Unique visitors 1.6M
  1. Apr
  2. May
  3. Jun
  4. Jul
  5. Aug
  6. Sep
Models by type

Models on Groq

ModelInputOutputvs officialContext
Qwen3.8 27B$0.80/1M$4/1M
from $0.50 at EmpirioLabs AI
+33%—
GPT OSS 120B$0.15/1M$0.60/1M
from $0.14 at Requesty
+329%—
GPT OSS 20B$0.075/1M$0.30/1M
from $0.09 at OpenRouter
+233%—
Canopy Labs Orpheus Arabic SaudiOnly here—$0.04/1K chars——
Canopy Labs Orpheus V1 EnglishOnly here—$0.022/1K chars——
Faster Whisper Large v3—$0.002/min
from $0.002 at Together AI
+23%—
GPT OSS Safeguard 20B$0.075/1M$0.30/1M
from $0.20 at Amazon Bedrock
+50%—
Llama Prompt Guard 2 22M$0.03/1M$0.03/1M
from $0.01 at Helicone
+200%—
Prompt Guard 2 86M$0.04/1M$0.04/1M
from $0.01 at Helicone
+300%—
Whisper Large v3 TurboCheapest—$0.000667/min——

Groq at a glance

10 models in total: 4 audio, 3 text, 3 moderation. The cheapest is Llama Prompt Guard 2 22M at $0.03 /1M. Qwen3.8 27B has the largest context window, 262K tokens. The most searched is Qwen3.8 27B, with about 120K Google searches a month. The newest is Qwen3.8 27B, released Aug 5, 2026.

Cheapest

  1. Llama Prompt Guard 2 22M$0.03/1M
  2. Prompt Guard 2 86M$0.04/1M
  3. GPT OSS 20B$0.30/1M
  4. GPT OSS Safeguard 20B$0.30/1M
  5. GPT OSS 120B$0.60/1M

Largest context

  1. Qwen3.8 27B262K
  2. GPT OSS 120B131K
  3. GPT OSS 20B131K
  4. GPT OSS Safeguard 20B131K
  5. Faster Whisper Large v30K

Most searched

  1. Qwen3.8 27B120K
  2. GPT OSS 120B35K
  3. GPT OSS 20B12K

Newest

  1. Qwen3.8 27BAug 5, 2026
  2. GPT OSS Safeguard 20BSep 18, 2025
  3. GPT OSS 120BAug 4, 2025
  4. GPT OSS 20BAug 4, 2025
  5. Llama Prompt Guard 2 22MApr 28, 2025