GPT-OSS 20B vs Llama 3.3 70B on Groq
Llama 3.3 70B on Groq was retired for self-serve accounts, so the practical question is what replaces it. This page puts GPT-OSS 20B on this endpoint next to the retired offer: price, limits and what changes in your code.
Groq figures are public list prices and documented limits, checked 2026-09-16 against the provider page. Inference APIs prices are read live from the price list.
- Cost: Groq listed it at a lower price, but it can no longer be called there, so the comparison that matters is between the options that still exist.
- Availability: Retired from Groq free and developer tiers on 2026-08-16; requests return 404 model_not_found. Enterprise contracts only.
Side by side
| GPT-OSS 20B | Llama 3.3 70B on Groq | |
|---|---|---|
| Served by | Inference APIs | Groq |
| Model author | OpenAI | OpenAI |
| Model id | openai/gpt-oss-20b | llama-3.3-70b-versatile |
| Input price | $0.04 / 1M tokens | $0.59 / 1M tokens |
| Output price | $0.19 / 1M tokens | $0.79 / 1M tokens |
| Context window | 131K tokens | 131K tokens |
| Capabilities | Chat, Reasoning, Tool calling, JSON mode | Chat, Tool calling, JSON mode |
| Weights | Open | Open |
| Status | Available | Retired |
| Limits | No per-minute or per-day request or token caps; usage draws on a prepaid balance | Retired from Groq free and developer tiers on 2026-08-16; requests return 404 model_not_found. Enterprise contracts only. |
What three workloads cost
List prices applied to the same work. The cheaper side of each row is in bold.
| Workload | GPT-OSS 20B | Llama 3.3 70B on Groq | Difference |
|---|---|---|---|
| Chat assistant — 10,000 turns of 800 tokens in, 300 out | $0.89 | $7.09 | 87% |
| Long prompts or RAG — 10,000 requests of 8,000 in, 500 out | $4.15 | $51.15 | 92% |
| Generation-heavy — 10,000 requests of 500 in, 2,000 out | $4.00 | $18.75 | 79% |
Groq listed it at a lower price, but it can no longer be called there, so the comparison that matters is between the options that still exist. Reasoning models bill their thinking as output tokens, so real output counts run higher than the visible answer; treat the generation-heavy row as a floor for them. The Llama 3.3 70B on Groq column is its last list price, shown for reference only.
When to choose which
- Llama 3.3 70B on Groq is no longer offered to self-serve accounts, so this is the one you can actually call
- Cost is the deciding factor; it is cheaper in every workload above
- You are replacing llama-3.1-8b-instant on Groq and want the small model Groq itself recommends
- Classification, extraction and short assistant replies where cost per call matters more than depth
Retired for self-serve accounts. There is no case for choosing it for new work.
Price, context and speed are measurable; answer quality on your task is not something a table can settle. Both take the same request, so the honest test is to run your own prompts through each.
Switching from Groq
Both speak the OpenAI wire format. The edit is the base URL, the key and the model id; then re-run your own prompts, because it is a different model.
import os
from openai import OpenAI
# before: Groq
# client = OpenAI(base_url="https://api.groq.com/openai/v1", api_key=os.environ["GROQ_API_KEY"])
# MODEL = "llama-3.3-70b-versatile"
# after: Inference APIs
client = OpenAI(base_url="https://api.inferenceapis.com/v1", api_key=os.environ["INFERENCE_API_KEY"])
MODEL = "openai/gpt-oss-20b"
resp = client.chat.completions.create(model=MODEL, messages=[{"role": "user", "content": "Hello"}], max_tokens=500)More detail, including Node, LangChain and LiteLLM: switching OpenAI-compatible providers.
FAQ
Which is cheaper, GPT-OSS 20B or Llama 3.3 70B on Groq?
Groq listed it at a lower price, but it can no longer be called there, so the comparison that matters is between the options that still exist. The table above uses list prices: GPT-OSS 20B $0.04 / 1M tokens input, $0.19 / 1M tokens output; Llama 3.3 70B on Groq $0.59 / 1M tokens input, $0.79 / 1M tokens output.
Can I still use Llama 3.3 70B on Groq?
Not on a self-serve account. Retired from Groq free and developer tiers on 2026-08-16; requests return 404 model_not_found. Enterprise contracts only.
Why does GPT-OSS 20B use more output tokens than the visible answer?
It is a reasoning model: it thinks before it answers and the thinking is billed as output tokens. Budget for that in output-heavy workloads, and set max_tokens to a few hundred or more so the answer is not cut off.
Spotted a price or limit that has changed? Tell us and we will re-check the provider page.