Meta

Llama 3.3 70B Instruct

Launched 2024-12-06 · Knowledge up to 2023-12 · Listed on OpenRouter 2024-12-06

Prices updated . Benchmarks updated .

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out). The Llama 3.3 instruction tuned text only model...

You send / 1M

$0.710

tokens in

It replies / 1M

$0.710

tokens out

Context

100K

tokens it can read at once

Typical request

$0.0013

1,000 words in, 400 out

Benchmarks

BenchmarkScoreKindDateSource
GPQA Diamond47.4%Independent2025-01-27Epoch AI
OTIS Mock AIME5.1%Independent2025-02-25Epoch AI
CritPt0.0%Independent2024-12-06CritPt, republished by Epoch AI
Arena Elo (text)1,274Independent2026-10-28LMArena

Data privacy

No training or retention: Yes. At least one provider serves it with zero data retention.

ProviderYou send / 1MIt replies / 1MUptimePrivacy policy
DeepInfra · fp8$0.100$0.32099%Read
Parasail · fp8$0.220$0.50099%Read
AkashML · fp8$0.200$0.520100%Read
CoreWeave · fp16$0.710$0.710100%Read
Google$0.720$0.720n/aRead
Google$0.720$0.720n/aRead
Groq$0.590$0.790100%Read
SambaNova$0.600$1.20100%Read
Cloudflare · fp8$0.293$2.2597%Read

Based on OpenRouter and Vercel AI Gateway data. Confirm terms with the provider before sending sensitive data.

What it can do

  • Reads: text
  • Writes: text
  • Weights: Open weights
  • Reasoning: No
  • Tool calling: Yes
  • Longest reply: 16K tokens
  • Repeated text is billed at $0.710 per 1M (prompt caching)

Compare with

Appears in

Estimate your monthly cost with Llama 3.3 70B Instruct

Sources and licenses

  • OpenRouter: Prices, context windows, modalities and the model list come from the OpenRouter public API. Data: OpenRouter.
  • Vercel AI Gateway: The no-training flag comes from the Vercel AI Gateway model list.
  • models.dev: Lab launch dates, open-weights flags and knowledge cutoffs come from models.dev. Copyright (c) 2025 models.dev, MIT License.
  • Epoch AI: Independent benchmark scores and the Epoch Capabilities Index (ECI) come from Epoch AI's Benchmarking Hub, used under CC BY 4.0. Epoch AI, Capabilities & benchmarking, epoch.ai/benchmarks. Scores Epoch marks as external keep their original license, and the source is named on each score.
  • LMArena: Arena Elo (text, webdev and vision) comes from the LMArena leaderboard dataset (lmarena-ai/leaderboard-dataset), CC BY 4.0, leaderboard date 2026-10-28.
  • Lab-reported scores, when an editor has added them, are the lab's own numbers. Each one links to the source and is labeled lab-reported. They are not used to rank the lists.
  • What a score measures: the benchmark guide.