404 model_not_found: llama-3.3-70b-versatile (Groq)

If your app started returning 404 The model `llama-3.3-70b-versatile` does not exist or you do not have access to it, the model was retired. Groq shut down llama-3.3-70b-versatile and llama-3.1-8b-instant on its free and developer tiers on 2026-08-16. Here is how to get your app working again.

Last verified September 16, 2026 against Groq deprecations · 5 min read

Provider	Groq
HTTP status	`404 Not Found`
Error code	`model_not_found`
Message	The model `llama-3.3-70b-versatile` does not exist or you do not have access to it.
When it happens	Since 2026-08-16, on any free or developer-tier request that names this model (or llama-3.1-8b-instant)
Can you wait it out?	No — the model was retired; the request will fail until you change the model or the provider

What happened

On 2026-08-16 Groq decommissioned llama-3.3-70b-versatile and llama-3.1-8b-instant for free and developer-tier usage. Any request pinned to those model ids now fails with a 404 model_not_found. Only enterprise committed-spend contracts were exempt.

Option 1 — switch to Groq's recommended replacement

Groq recommends migrating to openai/gpt-oss-120b (for the 70B) and openai/gpt-oss-20b (for the 8B). This is a one-line model-string change and keeps you on Groq's free tier — but note you inherit the same rate limits, and it is a different model with different behavior.

Example

- model="llama-3.3-70b-versatile"
+ model="openai/gpt-oss-120b"

Option 2 — keep running Llama 3.3 70B elsewhere

If your prompts, tests or output formatting were tuned for Llama 3.3 70B and you would rather not re-tune for a different model, the model itself is still served by other OpenAI-compatible providers. Point your base URL at one and keep the model:

Example

client = OpenAI(
    base_url="https://api.inferenceapis.com/v1",
    api_key=os.environ["INFERENCEAPIS_KEY"],
)
resp = client.chat.completions.create(
    model="llama-3.3-70b-versatile",   # still accepted here
    messages=[...],
)

Inference APIs accepts the exact llama-3.3-70b-versatile id so your code does not change beyond the base URL and key.

Where Inference APIs fits

Option 2 above keeps the same model on another provider. Inference APIs serves Llama 3.3 70B and accepts the exact id llama-3.3-70b-versatile, so the change is the base URL and key.

Create a free account API docs How a base-URL switch works

Something changed or wrong? Tell us and we will re-verify the entry.