Meta · Open weights
Llama 3.3 70b Instruct
Llama 3.3 is optimized for multilingual dialogue use cases and outperforms many of the available open source and closed chat models on common industry benchmarks.
- Tools
- Structured output
- Input
- Text
- Output
- Text
- Discovery tier
- Open weights
Pricing
USD per 1M tokens- Input
- $0.05
- Output
- $0.23
Reference prices as published by the model provider. API example
Specifications
- Context window
- 131.1K
- Max output
- 16.4K
- Average output (7 days)
- 37 tokens
- Longer than 2% of 465 models
Side by side
Llama 3.3 70b Instruct compared
Specifications, prices, speed and benchmark scores next to each other. Benchmarks below show every compared model too.
| Llama 3.3 70b InstructMetaThis model | Muse Spark OpenMetaRemove | |
|---|---|---|
| Provider | Meta | Meta |
| Discovery tier | Open weights | Frontier |
| Added | Feb 27, 2025 | Oct 15, 2026 |
| Limits | ||
| Context window | 131.07K | 1.05MBest |
| Max output | 16.38K | — |
| Average output | 37 tokens | 1.3K tokens |
| Pricing, USD per 1M tokens | ||
| Input | $0.05Best | $1.25 |
| Output | $0.23Best | $4.25 |
| Cache read | — | $0.15 |
| Performance | ||
| Throughput | 11.1 tok/s | — |
| Time to first token | 3.7 s | — |
| Request success (24h) | 99.9% | — |
| Benchmark indexes | ||
| Intelligence Index | 7.7 | 38.9Best |
| Coding Index | 11.9 | — |
| Math Index | 7.7 | — |
| Capabilities | ||
| Reasoning | No | Yes |
| Tools | Yes | Yes |
| Vision | No | Yes |
| Structured output | Yes | Yes |
| Reasoning effort | — | low, medium, high, xhigh |
Public evaluations
Benchmarks
Independent evaluation results, with how each score ranks among the models in the Noumen catalog.
- Llama 3.3 70b Instruct
- Muse Spark Open
Intelligence Index
- Llama 3.3 70b Instruct7.7
- Muse Spark Open38.9
Llama 3.3 70b Instruct: Better than 11% of 326 scored models
Indexes
Coding Index
Composite of the coding evaluations
- Llama 3.3 70b Instruct: 11.9
- Muse Spark Open: —
Math Index
Composite of the math evaluations
- Llama 3.3 70b Instruct: 7.7
- Muse Spark Open: —
Agentic work
AutomationBench-AA
Workflow automation, penalized for guardrail violations
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 49.7%
AutomationBench-AA tasks completed
Workflows finished end to end without a guardrail violation
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 20.9%
AA-Briefcasev1.1
Agentic knowledge work, rated head to head
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 1,279 Elo
GDPval-AAv2.1
Economically valuable occupational tasks, rated head to head
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 1,401 Elo
Document reasoning
GDP.pdf
Professional reasoning over PDF documents, all-pass rate
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 18.6%
AA-LCRv1.1
Reasoning across long documents
- Llama 3.3 70b Instruct: 15.7%
- Muse Spark Open: 78.3%Top
MLCR-AA
Long-context reasoning over medical records
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 7.8%
Reasoning and knowledge
Humanity’s Last Exam
Expert-level questions across academic fields
- Llama 3.3 70b Instruct: 3.6%
- Muse Spark Open: 43.9%Top
GPQA Diamond
Graduate-level science questions
- Llama 3.3 70b Instruct: 49.8%
- Muse Spark Open: 89.0%Top
MMLU-Pro
Multi-subject knowledge and reasoning
- Llama 3.3 70b Instruct: 71.3%
- Muse Spark Open: —
IFBench
Following precise output instructions
- Llama 3.3 70b Instruct: 47.1%
- Muse Spark Open: —
Coding
Terminal-Benchv4.0
Practical software engineering in a terminal
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 8.1%
SciCode
Scientific computing in Python
- Llama 3.3 70b Instruct: —
- Muse Spark Open: 58.4%
LiveCodeBench
Recent competitive programming problems
- Llama 3.3 70b Instruct: 28.8%
- Muse Spark Open: —
Math
AIME 2025
Competition mathematics, 2025 problems
- Llama 3.3 70b Instruct: 7.7%
- Muse Spark Open: —
AIME
Competition mathematics
- Llama 3.3 70b Instruct: 30.0%
- Muse Spark Open: —
MATH-500
Competition math problems
- Llama 3.3 70b Instruct: 77.3%
- Muse Spark Open: —
Legacy evaluations
Terminal-Bench Hard
Earlier terminal task suite
- Llama 3.3 70b Instruct: 3.0%
- Muse Spark Open: —
τ²-Bench Telecom
Tool use in a customer-support simulation
- Llama 3.3 70b Instruct: 26.6%
- Muse Spark Open: —
AA-LCR (unversioned)
Long-context reasoning, earlier grading
- Llama 3.3 70b Instruct: 15.7%
- Muse Spark Open: 78.3%Top
Scores from Artificial Analysis, updated Oct 26, 2026. Artificial Analysis
- Llama 3.3 70b Instruct
- Muse Spark Open
| Leaderboard | Llama 3.3 70b Instruct | Muse Spark Open |
|---|---|---|
| Website | Not in top 20 | #15 · 1306 |
Ranks among Noumen catalog models on Design Arena leaderboards. Only top-20 placements are listed.Design Arena
Live routing data
Performance and reliability
Recent averages across requests routed for this model. Your own speed depends on prompt length, reasoning effort and load.
- Throughput
- 11.1 tok/s
- Faster than 1% of 247 models
- Time to first token
- 3.7 s
- Quicker than 46% of 247 models
- Prompt cache hits
- 0.0%
- Request success (24h)
- 99.9%
- Reliable
Cache hits are the share of input tokens recently served from a prompt cache across all requests. They are not a forecast of your own hit rate.
- Healthy (98% or more)
- Degraded (90–98%)
- Down (under 90%)
- No requests
Developers
Use it through the API
Noumen’s API is OpenAI-compatible: any Chat Completions client works with a base URL and a key.
curl https://www.noumen.xyz/api/v1/chat/completions \
-H "Authorization: Bearer $NOUMEN_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "meta-llama/llama-3.3-70b-instruct",
"messages": [{ "role": "user", "content": "Explain this stack trace." }]
}'import os
from openai import OpenAI
client = OpenAI(base_url="https://www.noumen.xyz/api/v1", api_key=os.environ["NOUMEN_API_KEY"])
reply = client.chat.completions.create(
model="meta-llama/llama-3.3-70b-instruct",
messages=[{"role": "user", "content": "Explain this stack trace."}],
)
print(reply.choices[0].message.content)Meta
Related models
Other Meta models, starting with the same family.