AI API price index · open-model host · read 16 Oct 2026
Groq API pricing
13 prices as Groq publishes them — 9 text models, 2 speech to text, 2 text to speech — read on its own pricing page on 16 Oct 2026. Its current text models cost $0.30 to $4.00 per million output tokens.
Free tier A Free plan exists: the rate-limits page has a 'Free Plan Limits' tab (e.g. openai/gpt-oss-120b 30 RPM, 1K RPD, 8K TPM, 200K TPD; whisper-large-v3 20 RPM, 7.2K ASH). Upgrading to the pay-as-you-go Developer tier needs a payment method; 'there's no immediate charge'. Billing FAQ: 'We occasionally offer promotional credits, such as during hackathons and special events.' No free dollar credit is stated.
What to know before you build on Groq
- groq.com/pricing now redirects to the homepage; the price table is on console.groq.com/docs/models.
- Llama 3.1 8B and Llama 3.3 70B are tagged Enterprise with "Contact Sales" — Groq publishes no per-token price for them any more.
- Batch takes 50% off. Cached input is 50% off on the GPT OSS models, and the two discounts do not stack.
- A Free plan exists, defined by rate limits only — for GPT OSS 120B, 30 requests a minute and 1,000 a day.
- Several models are marked Preview: "evaluation purposes only" and may be withdrawn at short notice.
Against the other ten
Across the index, 85 current text models run from $0.10 to $50.00 per million output tokens — 500×. Groq’s cheapest, GPT OSS 20B (openai/gpt-oss-20b) at $0.30, has 9 cheaper text models elsewhere; its dearest, Qwen/Qwen3.8-27B (qwen/qwen3.8-27b), costs $4.00. A cheaper model is a different model, not the same one for less — compare capability before price.
Speech to text across the index runs from $0.04 to $0.552 per audio hour; Groq’s Whisper Large V3 Turbo (whisper-large-v3-turbo) is $0.04.
Text to speech across the index runs from $15.00 to $80.00 per million characters; Groq’s cheapest ranked voice, Canopy Labs Orpheus V1 English (canopylabs/orpheus-v1-english), is $22.00.
The same model elsewhere
Open-weight models Groq sells that other hosts price too — the only like-for-like comparison in the index.
- Whisper Large v3 — Together AI $0.09/h · Groq $0.111/h — 1.2× apart
- gpt-oss-120b — Fireworks AI $0.15 / $0.60 · Groq $0.15 / $0.60 · Together AI $0.15 / $0.60 — the same everywhere
Input / output per 1M tokens, standard tier.
Text models
| Model | Input | Cached input | Output | Note |
|---|---|---|---|---|
| GPT OSS 120B (openai/gpt-oss-120b) | $0.15 | $0.075 | $0.60 | — |
| GPT OSS 20B (openai/gpt-oss-20b) | $0.075 | $0.037 | $0.30 | — |
| Safety GPT OSS 20B (openai/gpt-oss-safeguard-20b) | $0.075 | $0.037 | $0.30 | Preview model (evaluation only) · Not ranked: a safety classifier, not a general model. |
| Qwen/Qwen3.8-27B (qwen/qwen3.8-27b) | $0.80 | — | $4.00 | Preview model (evaluation only); text + image input, each image counts as 2048 input tokens |
| Llama Prompt Guard 2 22M (meta-llama/llama-prompt-guard-2-22m) | $0.03 | — | $0.03 | Preview model (evaluation only) · Not ranked: a safety classifier, not a general model. |
| Prompt Guard 2 86M (meta-llama/llama-prompt-guard-2-86m) | $0.04 | — | $0.04 | Preview model (evaluation only) · Not ranked: a safety classifier, not a general model. |
| Llama 3.1 8B (llama-3.1-8b-instant)enterprise | — | — | — | No public price: tagged Enterprise, price column reads Contact Sales · Not ranked: no public price — Contact Sales. |
| Llama 3.3 70B (llama-3.3-70b-versatile)enterprise | — | — | — | No public price: tagged Enterprise, price column reads Contact Sales · Not ranked: no public price — Contact Sales. |
| MiniMax M2.7 (minimaxai/minimax-m2.7)enterprise | — | — | — | Preview model; no public price: tagged Enterprise, Contact Sales · Not ranked: no public price — Contact Sales. |
Speech to text
| Model | Price | Note |
|---|---|---|
| Whisper (whisper-large-v3) | $0.111 per hour | — |
| Whisper Large V3 Turbo (whisper-large-v3-turbo) | $0.04 per hour | — |
Text to speech
| Model | Price | Note |
|---|---|---|
| Canopy Labs Orpheus V1 English (canopylabs/orpheus-v1-english) | $22.00 per 1M characters | Preview model; unit is per 1M characters as printed (= $0.022 per 1K characters) |
| Canopy Labs Orpheus Arabic Saudi (canopylabs/orpheus-arabic-saudi) | $40.00 per 1M characters | Preview model; unit is per 1M characters as printed (= $0.04 per 1K characters) |
7 of these 13 prices are ranked in the index; the rest say why in their note. Text and embeddings per 1M tokens unless the row says otherwise.
Discounts and surcharges, as published
- Batch API: 'Pricing is at a 50% cost discount compared to synchronous API pricing.' Batch models listed: openai/gpt-oss-20b, openai/gpt-oss-120b, llama-3.3-70b-versatile, llama-3.1-8b-instant, meta-llama/llama-guard-4-12b, whisper-large-v3, whisper-large-v3-turbo. 24-hour to 7-day processing window.
- Prompt caching: 'There is a 50% discount for cached input tokens.' Supported only on openai/gpt-oss-20b, openai/gpt-oss-120b, openai/gpt-oss-safeguard-20b. 'The batch discount does not stack with prompt caching discounts.'
- Flex service tier: 'Pricing matches the on-demand tier' (10x higher rate limits, paid customers only, may fail with 498 capacity_exceeded).
- Performance tier (enterprise only): provisioned throughput bundles, not per-token; 'Reach out to inquire about pricing.'
Where these numbers come from
Read on console.groq.com/docs/models on 16 Oct 2026, checked against 14 more of Groq’s own pages. Macrostack measures nothing and is paid by no provider on this index. Prices change; the date is the claim.
The other providers in the index
- The AI API price index — every provider, cheapest first, and the download
- OpenAI API pricing — model maker
- Anthropic API pricing — model maker
- Google Gemini API pricing — model maker
- Mistral AI API pricing — model maker
- Cohere API pricing — model maker
- Together AI API pricing — open-model host
- Fireworks AI API pricing — open-model host
- Replicate API pricing — open-model host
- Deepgram API pricing — voice
- ElevenLabs API pricing — voice
Questions
- How much does the Groq API cost?
- On 16 Oct 2026, Groq's current text models cost $0.30 to $4.00 per million output tokens at the standard tier (GPT OSS 20B (openai/gpt-oss-20b) to Qwen/Qwen3.8-27B (qwen/qwen3.8-27b)). The full list, with input and cached-input prices, is on this page.
- Does Groq have a free tier?
- What Groq's own pages say: A Free plan exists: the rate-limits page has a 'Free Plan Limits' tab (e.g. openai/gpt-oss-120b 30 RPM, 1K RPD, 8K TPM, 200K TPD; whisper-large-v3 20 RPM, 7.2K ASH). Upgrading to the pay-as-you-go Developer tier needs a payment method; 'there's no immediate charge'. Billing FAQ: 'We occasionally offer promotional credits, such as during hackathons and special events.' No free dollar credit is stated.
- Does Groq offer discounts?
- Batch API: 'Pricing is at a 50% cost discount compared to synchronous API pricing.' Batch models listed: openai/gpt-oss-20b, openai/gpt-oss-120b, llama-3.3-70b-versatile, llama-3.1-8b-instant, meta-llama/llama-guard-4-12b, whisper-large-v3, whisper-large-v3-turbo. 24-hour to 7-day processing window. Prompt caching: 'There is a 50% discount for cached input tokens.' Supported only on openai/gpt-oss-20b, openai/gpt-oss-120b, openai/gpt-oss-safeguard-20b. 'The batch discount does not stack with prompt caching discounts.' Flex service tier: 'Pricing matches the on-demand tier' (10x higher rate limits, paid customers only, may fail with 498 capacity_exceeded). Performance tier (enterprise only): provisioned throughput bundles, not per-token; 'Reach out to inquire about pricing.'