Text models
Arena rating
THE MODEL, IN DETAIL / Meta
llama-3.3-70b-instruct
meta-llama/llama-3.3-70b-instructOpen weightsThe 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...
#276
Context window
16.4K max output tokens
Input price
USD / million input tokens
Output price
USD / million output tokens
Inference providers
11 provider routes observed
Measured output speed
No model-level measurement
01 / PERFORMANCE
A clearer view of capability.
Read the score, understand the context, and explore where this model sits among its peers.
Performance fingerprint
Individual published accuracy scores · not a composite index
At least three published accuracy scores are needed for this view.
Performance meets price
Text models · same published source and version
GoogleAnthropicMetaxiaomi
In good company
Text models · Arena · official dataset
416 models in the source cohort. Showing selected nearby models. Evaluation settings can differ.
Evidence, by task.
Coverage is not a capability score.
0published results
0 evaluation sources
0 evaluation sources
ReasoningAwaiting evidence
CodingAwaiting evidence
MathematicsAwaiting evidence
VisionAwaiting evidence
AgenticAwaiting evidence
Long contextAwaiting evidence
Documented features, task evaluations, and preference rankings describe different things. Missing coverage does not imply a model cannot perform a task.
Explore our methodologyWhere people rank it
Published rankings for this exact model or configuration. Each source keeps its original metric.
02 / BENCHMARK EXPLORER
The numbers behind the model.
Original results, exact versions, and a direct path to the evidence behind every score.
Percentage-based evaluations will appear here when published.
0 percentage-based evaluations. Each benchmark measures a different task; compare results within the same evaluation and protocol.
Artificial Analysis indices
Published in the OpenRouter catalog · source-defined index values
Catalog observation: 17 Oct 2026. Index versions and evaluation settings are not supplied in this feed. These are not Singularity composite scores.
Artificial Analysis03 / COST & VALUE
Know what your workload costs.
Explore token prices, see where the spend goes, and model a workload of your own.
Input$0.1per million tokens
Output$0.32per million tokens
Cache read—per million tokens
Cache write—per million tokens
Workload cost calculator
Change the numbers. See the trade-offs.
Cache-read pricing is not documented for this selection.
Token estimate only. Excludes cache writes, tools, and other provider fees. Actual token usage and provider limits vary.
Your estimated spend
1,000 requests · 11M tokens
Uncached input$1
Cached input$0
Output$0.32
Per request$0.0013
Without cache discount$1.32
Estimated cache savings$0
04 / INFERENCE NETWORK
One model. Different ways to run it.
Compare inference routes by price, context, quantization, and reported availability.
| Provider / route | Input / M | Output / M | Cache / M | Context | Output limit | Speed | Latency | Availability · 1d | Rates |
|---|---|---|---|---|---|---|---|---|---|
| DDeepInfraFP8 | $0.1 | $0.32 | — | 131.1K | 16.4K | — | — | 100% | |
| NNovitaBF16 | $0.135 | $0.4 | — | 12.3K | 11.1K | — | — | 99.79% | |
| AAkashMLFP8 | $0.2 | $0.52 | $0.1 | 131.1K | 128K | — | — | 99.85% | |
| PParasailFP8 | $0.22 | $0.5 | $0.11 | 131.1K | 16.4K | — | — | 100% | |
| CCloudflareFP8 | $0.293 | $2.253 | — | 24K | 21.6K | — | — | 99.69% | |
| SSambaNovaQuantization not disclosed | $0.45 | $0.9 | — | 131.1K | 3.1K | — | — | 99.78% | |
| GGroqQuantization not disclosed | $0.59 | $0.79 | $0.295 | 131.1K | 32.8K | — | — | 100% | |
| CCoreWeaveFP16 | $0.71 | $0.71 | $0.71 | 128K | 115.2K | — | — | 100% | |
| GGoogleQuantization not disclosed | $0.72 | $0.72 | — | 128K | 8.2K | — | — | — | |
| GGoogleQuantization not disclosed | $0.72 | $0.72 | — | 128K | 115.2K | — | — | — | |
| TTogetherQuantization not disclosed | $1.04 | $1.04 | — | 131.1K | 2K | — | — | 92.34% |
Snapshot observed 17 Oct 2026. Speed and latency are the reported 30-minute median when available. This is a dated snapshot, not live monitoring.
Provider price comparison
USD per million tokens · input and output rates
InputOutput
DeepInfrafp8
Input
$0.1
Output
$0.32
Novitabf16
Input
$0.135
Output
$0.4
AkashMLfp8
Input
$0.2
Output
$0.52
Parasailfp8
Input
$0.22
Output
$0.5
Cloudflarefp8
Input
$0.293
Output
$2.253
SambaNova
Input
$0.45
Output
$0.9
Groq
Input
$0.59
Output
$0.79
CoreWeavefp16
Input
$0.71
Output
$0.71
Google
Input
$0.72
Output
$0.72
Google
Input
$0.72
Output
$0.72
Reported availability
Provider availability over the selected observation window
Output speed
Reported 30-minute p50 · tokens per second
No provider throughput measurements were returned in this snapshot.
Response latency
Reported 30-minute p50 · seconds
No provider latency measurements were returned in this snapshot.
Context by provider
Actual route limits can differ from the catalog maximum
05 / CAPABILITIES & MODALITIES
What goes in. What comes out.
A map of documented interfaces and API features, separate from benchmark performance.
INPUT MODALITIES
Text
llama-3.3-70b-instructModel interface
OUTPUT MODALITIES
Text
Feature support
Listed catalog API parameters and documented pricing features
ReasoningNot listed
Tool callingSupported
Tool choiceSupported
Structured outputsSupported
Response formatSupported
Reasoning detailsNot listed
Seed controlSupported
Stop sequencesSupported
Prompt cachingNot documented
TemperatureSupported
Room to work
Context capacity and completion limits
131.1Ktotal context tokens
Context window
16.4Kmax output
Output tokens share the total context budget. The maximum completion length is a separate limit, and individual provider routes may have smaller limits.
- Catalog context
- 131,072 tokens
- Maximum output
- 16,384 tokens
- Top-provider context
- 131,072 tokens
Features vary by route
Check the provider before depending on a specific API feature.
| Provider route | Reasoning | Tools | Structured output | Response format | Implicit caching |
|---|---|---|---|---|---|
| DeepInfra · FP8 | Not listed | Supported | Supported | Supported | Not listed |
| Novita · BF16 | Not listed | Supported | Not listed | Not listed | Not listed |
| AkashML · FP8 | Not listed | Supported | Supported | Supported | Supported |
| Parasail · FP8 | Not listed | Not listed | Supported | Supported | Supported |
| Cloudflare · FP8 | Not listed | Not listed | Not listed | Supported | Not listed |
| SambaNova | Not listed | Not listed | Supported | Supported | Not listed |
| Groq | Not listed | Supported | Not listed | Supported | Supported |
| CoreWeave · FP16 | Not listed | Supported | Supported | Supported | Supported |
| Not listed | Supported | Supported | Supported | Not listed | |
| Not listed | Not listed | Not listed | Supported | Not listed | |
| Together | Not listed | Supported | Supported | Supported | Not listed |
06 / UNDER THE HOOD
The complete specification sheet.
Exact identities, model architecture, limits, and the defaults that shape a request.
Identity & architecture
- Developer
- Meta
- Exact identity
llama-3.3-70b-instruct- Subject
- Model Configuration
- Access
- Open weights
- License
- Llama-3.3
- Parameters
- Not disclosed
- Tokenizer
- Llama3
- Instruction format
- llama3
Reasoning & controls
- Reasoning
- Not listed
- Mandatory reasoning
- Not documented
- Reasoning by default
- Not documented
- Published configuration
- Not disclosed
- Supported reasoning efforts
- Not documented
- Default reasoning effort
- Not disclosed
- API parameters
- 17
- Provider moderation
- No
- Quantization variants
- fp8, bf16, fp16
Dates & limits
- Catalog added
- 6 Dec 2024
- Knowledge cutoff
- 2023-12-31
- Catalog expiration
- Not documented
- Specifications observed
- 17 Oct 2026
- Providers observed
- 17 Oct 2026
- Request limits
- Provider-specific / not supplied
- Canonical catalog ID
meta-llama/llama-3.3-70b-instruct
API parameters & defaults
Availability is catalog-wide. Defaults can vary across providers.
frequency_penaltySupported · default not suppliedlogit_biasSupported · default not suppliedlogprobsSupported · default not suppliedmax_tokensSupported · default not suppliedmin_pSupported · default not suppliedpresence_penaltySupported · default not suppliedrepetition_penaltySupported · default not suppliedresponse_formatSupported · default not suppliedseedSupported · default not suppliedstopSupported · default not suppliedstructured_outputsSupported · default not suppliedtemperatureSupported · default not suppliedtool_choiceSupported · default not suppliedtoolsSupported · default not suppliedtop_kSupported · default not suppliedtop_logprobsSupported · default not suppliedtop_pSupported · default not supplied07 / BUILD WITH THIS MODEL
From exploration to your first request.
Use the documented catalog identifier and keep your API key in an environment variable.
Make an API request
const response = await fetch(
"https://openrouter.ai/api/v1/chat/completions",
{
method: "POST",
headers: {
Authorization: `Bearer ${process.env.OPENROUTER_API_KEY}`,
"Content-Type": "application/json",
},
body: JSON.stringify({
"model": "meta-llama/llama-3.3-70b-instruct",
"messages": [
{
"role": "user",
"content": "Explain the trade-offs in this design."
}
]
}),
},
);
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const data = await response.json();
console.log(data.choices[0].message.content);08 / PROVENANCE & HISTORY
Every number has a story.
Follow the source, check the observation date, and distinguish published results from model specifications.
The source register
Direct links to the underlying records
OpenRouter model catalog
02Architecture, modalities, parameters, and starting prices
Observed 17 Oct 2026OpenRouter provider routes
03Provider limits, prices, availability, and reported measurements
Observed 17 Oct 2026Text models ranking
Arena rating · source-reported ranking
Observed 15 Oct 2026The observation trail
Recorded events, in chronological order
Added to the model catalog
Catalog creation date, as recorded by OpenRouter.
Model specifications observed
Catalog pricing, limits, and supported parameters captured.
Provider routes observed
11 inference routes in this snapshot.
Observation dates describe when data was captured. They do not represent newly run evaluations or continuous performance monitoring.
Published observation ledger
Historical evidence, with source and version preserved.
No published benchmark observations for this selection. Published preference rankings appear separately above.
09 / A FEW MORE DETAILS
Questions worth asking.
Quick answers, drawn from this model’s connected specifications.
How much does llama-3.3-70b-instruct cost?
Catalog starting rates are $0.1 per million input tokens and $0.32 per million output tokens. Compare provider routes and service tiers above for their specific prices.
How much context can it handle?
131,072 tokens in the catalog context window. The catalog lists a maximum of 16,384 completion tokens. Provider limits can differ.
Does it support tools and structured output?
Tool calling: listed as supported. Structured outputs: listed as supported. See the provider feature matrix for route-specific differences.
How should I interpret these benchmark results?
Read the benchmark name, exact version, metric, source, and model configuration together. Different evaluation settings can change the result. Preference ratings and source-defined composite indices remain separate from task accuracy.
Are these prices and provider measurements live?
This page uses dated source snapshots. Catalog data was observed on 17 Oct 2026. Provider routes were observed on 17 Oct 2026. Missing measurements are not estimated or treated as zero.
KEEP EXPLORING
More from Meta.
Explore related identities from the same developer, or take them into a side-by-side comparison.