Skip to content
Not available in this workspace
OpenRouterOpenRouter
© 2026 OpenRouter, Inc

Product

  • Chat
  • Rankings
  • Benchmarks
  • Apps
  • Discover
  • Models
  • Collections
  • Providers
  • Pricing
  • Enterprise
  • Labs

Company

  • About
  • Blog
  • Careers
    Hiring
  • Privacy
  • Terms of Service
  • Trust Center
  • Support
  • Works With OR
  • Data

Developer

  • Documentation
  • API Reference
  • Developer Platform
  • Status

Connect

  • Discord
  • GitHub
  • LinkedIn
  • X
  • YouTube
Favicon for meta-llama

Meta: Llama 3.3 70B Instruct

meta-llama/llama-3.3-70b-instruct

Model weights
Compare

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 is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.

Supported languages: English, German, French, Italian, Portuguese, Hindi, Spanish, and Thai.

Model Card(opens in new tab)

Modalities

In / Out Price

$0.10 / $0.32per 1M

Context

131K

Released

Dec 6, 2024

Knowledge Cutoff

Dec 2023

Compare
ProvidersPricingPerformanceUptimeBenchmarksAppsActivityFAQExplore

Providers

Different companies host the same model. OpenRouter routes your request to one of them based on the routing mode you pick — Balanced (price + speed), Nitro (fastest), or Exacto (highest tool-calling accuracy).

Pricing

The average price customers actually pay for this model, next to the prices providers post. Caching and discounts mean the price actually paid is often well below the listed one.

Performance

Throughput is how fast the model writes (tokens per second — higher is better). Latency is total round-trip time (lower is better). TTFT is time-to-first-token — how long before you see anything appear (lower is better).

Uptime

Uptime is the percentage of the past 3 days that at least one provider was responding to requests. Availability is the percentage of time that inference was successfully served. OpenRouter continuously monitors and uses the next-best provider when one returns an error.

Benchmarks

Scores on standardized evaluations. Higher percentages are better — and rank percentile shows where this model lands among all models on OpenRouter.

Apps

Public apps that send the most traffic to this model. Good signal for what real production workloads look like — and a hint at which use cases this model is best suited for.

Activity

Token volume and request traffic to this model over time.

Quick Start

Drop-in code to call this model. OpenRouter's API is OpenAI-compatible — most SDKs work by just swapping the base URL. The only thing that changes between models is the model slug below.

Explore more models

AI Model RankingsRanking
$0.10$0.32--0.54s11 tps
97.94%
$0.20$0.52$0.100.69s19 tps
99.56%
$0.22$0.50$0.110.64s20 tps
99.45%
$0.25$0.75$0.130.40s50 tps
99.67%
$0.293$2.253--0.60s30 tps
98.67%
25% off
$0.60$0.45$1.20$0.90--0.68s59 tps
98.28%
$0.59$0.79$0.2950.28s15 tps
99.73%
$0.71$0.71$0.710.29s61 tps
99.98%
$0.72$0.72--0.34s42 tps
99.96%
$0.72$0.72--0.31s23 tps
89.73%
$1.04$1.04--0.64s40 tps
98.78%
$0.13$0.40--3.08s5 tps
58.01%
$0.135$0.40--0.85s23 tps
88.51%

Throughput

61tok/s

P50, best across providers

Latency

0.28s

P50, best provider

AutoExacto Benchmarks
GPQA DiamondTAU-Bench
Groq
50.5%--
SambaNova Turbo
48.3%--
Parasail
47.5%--
Crusoe
46.6%--
NovitaAI
47.0%40.8%
+8 more providers
Uptime (3d)

100.00%

Availability (3d)

99.93%

Availability over the last 3 days

Last 72 hours
Availability 99.93%
3 Days Ago2 Days AgoYesterdayNow

Availability over the last 24 hours

OpenRouter Availability
99.95%
Without Routing
93.88%

When an error occurs in an upstream provider, we can recover by routing to another healthy provider, if your request filters allow it. You can access per-provider uptime data programmatically through the Endpoints API. Learn more about our load balancing and customization options.

1.
Favicon for https://dialos.mediaha.us/
DialOS Ask Merlin
new
8.4Btokens
2.
Favicon for https://esmed2021.org/
ESMED Email App V3
new
3.34Btokens
3.
Favicon for https://fvchat.app/
FVChat
new
2.82Btokens
4.
Favicon for https://kymai.ai/
KYM AI Apsara - Hailey Cooper
new
1.74Btokens
5.
Favicon for https://shapes.inc/
shapes inc
General purpose social agents
1.45Btokens

Frequently asked questions

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 is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.

Llama 3.3 70B Instruct costs $0.10/M input tokens and $0.32/M output tokens.

Llama 3.3 70B Instruct has a 131,072 token context window. It supports up to 16,384 completion tokens.

Yes. Llama 3.3 70B Instruct accepts tools and tool_choice for function calling. It supports response_format for JSON output, without JSON-schema enforcement.

Llama 3.3 70B Instruct is served by 12 providers on OpenRouter: DeepInfra (Turbo), Nebius Token Factory, NovitaAI, AkashML, Parasail, Crusoe, Cloudflare, SambaNova Turbo and 4 more. Requests are routed to the best available provider, with automatic failover to the others, and you can pin or exclude providers with provider routing.

Llama 3.3 70B Instruct was released on December 6, 2024. Its knowledge cutoff is December 31, 2023.