Groq pricing and rate limits
Models and cost at your volume
GPT OSS 120B: $0.27 / mo base $0 + Input tokens overage $0.15 + Output tokens overage $0.12
Uses the shared Overage cost function. Example volumes you can edit, not vendor data. Cached, batch and cache-write prices are listed below but not added to this total.
| GPT OSS 120B | 131K | $0.15openai/gpt-oss-120b | 500 | $0.15 input$0.60 output | 250K TPM1K RPM | 131,072 | 65,536source | $0.60openai/gpt-oss-120b | 500 | $0.15 input$0.60 output | 250K TPM1K RPM | 131,072 | 65,536source | unlisted | unlisted | unlisted | unlisted | $0.27 |
| GPT OSS 20B | 131K | $0.075openai/gpt-oss-20b | 1000 | $0.075 input$0.30 output | 250K TPM1K RPM | 131,072 | 65,536source | $0.30openai/gpt-oss-20b | 1000 | $0.075 input$0.30 output | 250K TPM1K RPM | 131,072 | 65,536source | unlisted | unlisted | unlisted | unlisted | $0.135 |
| Llama Prompt Guard 2 22M | 512 | $0.03meta-llama/llama-prompt-guard-2-22m | - | $0.03 input$0.03 output | 30K TPM100 RPM | 512 | 512source | $0.03meta-llama/llama-prompt-guard-2-22m | - | $0.03 input$0.03 output | 30K TPM100 RPM | 512 | 512source | unlisted | unlisted | unlisted | unlisted | $0.036 |
| Prompt Guard 2 86M | 512 | $0.04meta-llama/llama-prompt-guard-2-86m | - | $0.04 input$0.04 output | 30K TPM100 RPM | 512 | 512source | $0.04meta-llama/llama-prompt-guard-2-86m | - | $0.04 input$0.04 output | 30K TPM100 RPM | 512 | 512source | unlisted | unlisted | unlisted | unlisted | $0.048 |
| Safety GPT OSS 20B | 131K | $0.075openai/gpt-oss-safeguard-20b | 1000 | $0.075 input$0.30 output | 150K TPM1K RPM | 131,072 | 65,536source | $0.30openai/gpt-oss-safeguard-20b | 1000 | $0.075 input$0.30 output | 150K TPM1K RPM | 131,072 | 65,536source | unlisted | unlisted | unlisted | unlisted | $0.135 |
| Qwen/Qwen3.8-27B | 131K | $0.80qwen/qwen3.8-27b | 450 | $0.80 input$4.00 output | 250K TPM1K RPM | 131,072 | 16,384 | 20 MB |source | $4qwen/qwen3.8-27b | 450 | $0.80 input$4.00 output | 250K TPM1K RPM | 131,072 | 16,384 | 20 MB |source | unlisted | unlisted | unlisted | unlisted | $1.60 |
USD per 1M tokens. "+" marks a model whose price changes with volume; the standard band is shown. Context in tokens (6 models).
Prices in USD, per the vendor's billing period. Hover or focus a number for the line it came from, with a link to the stored snapshot. "unlisted" means the vendor does not publish it; we never fill the gap.
Rate limits
| Tier | Metric | Limit | Window | Scope | How we know |
|---|---|---|---|---|---|
| GPT OSS 120B | API requests | 1,000openai/gpt-oss-120b | 500 | $0.15 input$0.60 output | 250K TPM1K RPM | 131,072source | per minute | per key | documented |
| GPT OSS 20B | API requests | 1,000openai/gpt-oss-20b | 1000 | $0.075 input$0.30 output | 250K TPM1K RPM | 131,072source | per minute | per key | documented |
| Llama Prompt Guard 2 22M | API requests | 100meta-llama/llama-prompt-guard-2-22m | - | $0.03 input$0.03 output | 30K TPM100 RPM | 512source | per minute | per key | documented |
| Prompt Guard 2 86M | API requests | 100meta-llama/llama-prompt-guard-2-86m | - | $0.04 input$0.04 output | 30K TPM100 RPM | 512source | per minute | per key | documented |
| Safety GPT OSS 20B | API requests | 1,000openai/gpt-oss-safeguard-20b | 1000 | $0.075 input$0.30 output | 150K TPM1K RPM | 131,072source | per minute | per key | documented |
| Qwen/Qwen3.8-27B | API requests | 1,000qwen/qwen3.8-27b | 450 | $0.80 input$4.00 output | 250K TPM1K RPM | 131,072source | per minute | per key | documented |
Change timeline
No changes recorded.
Tracking began 2026-10-24; no change has been detected since.
Sources
| URL | Kind | Last captured | Snapshot |
|---|---|---|---|
| console.groq.com/docs/models | pricing | 2026-10-24 | snap_groq-pricing-0_20261023T173956Z |
| console.groq.com/docs/rate-limits | limits | 2026-10-24 | snap_groq-limits-0_20261023T173958Z |
Report a correction
Spotted a wrong or missing number on Groq? Send the field, the value you see and a source link. Reports with a source are reviewed first.