Groq is a cloud platform that resells models with its own infrastructure, billing and regional hosting. Its prices tend to track the first-party rate, with the platform's own quotas and service-level terms attached.
It lists 16 models on this page. 9 of them have a public price. 1 is offered at $0. Coverage spans text and chat (9), speech and audio (4), and classification and safety (3). The lowest-priced model hosted here is Llama Prompt Guard 2 22M.
| Cheapest here? | |||||
|---|---|---|---|---|---|
| ALLaM-2-7ballam-2-7b | Free | — | 4,096 | Only channel | |
| Llama Prompt Guard 2 22MClassify / Safetymeta-llama/llama-prompt-guard-2-22m | $0.03 | $0.03 | — | 512 | 3.0× pricier |
| Meta Llama Prompt Guard 2 86MClassify / Safetymeta-llama/llama-prompt-guard-2-86m | $0.04 | $0.04 | — | 512 | 4.0× pricier |
| Llama 3.1 8Bllama-3.1-8b-instant | $0.05 | $0.08 | — | 131,072 | 1.0× pricier |
| GPT OSS 20Bopenai/gpt-oss-20b | $0.075 | $0.30 | $0.038 | 131,072 | 3.6× pricier |
| GPT OSS Safeguard 20BClassify / Safetyopenai/gpt-oss-safeguard-20b | $0.075 | $0.30 | — | 131,072 | 1.3× pricier |
| GPT OSS 120Bopenai/gpt-oss-120b | $0.15 | $0.60 | $0.075 | 131,072 | 4.4× pricier |
| Llama 3.3 70Bllama-3.3-70b-versatile | $0.59 | $0.79 | — | 131,072 | 1.1× pricier |
| Qwen3.6 27Bqwen/qwen3.6-27b | $0.60 | $3.00 | $0.30 | 131,072 | 2.3× pricier |
| Qwen3.8 27Bqwen/qwen3.8-27b | $0.80 | $4.00 | — | 131,042 | 14.1× pricier |
| Whisper Large v3 TurboSpeech / Audiowhisper-large-v3-turbo | — | — | — | — | Only channel |
| Whisper 3 LargeSpeech / Audiowhisper-large-v3 | — | — | — | — | — |
| Canopy Labs Orpheus Arabic SaudiSpeech / Audiocanopylabs/orpheus-arabic-saudi | — | — | — | 4,000 | Only channel |
| Canopy Labs Orpheus V1 EnglishSpeech / Audiocanopylabs/orpheus-v1-english | — | — | — | 4,000 | Only channel |
| Compoundgroq/compound | — | — | — | 131,072 | Only channel |
| Compound Minigroq/compound-mini | — | — | — | 131,072 | Only channel |