Llama-3.3-70B-Instruct Meta · Released Dec 6, 2024 · Open weights
Latest Llama-3.3-70B-Instruct Popular open Llama workhorse for multilingual chat, coding, and self-hosting
Strongest at Science Better than 0% of the past year's models · one benchmark so far
Price per 1M tokens $0.1 / $0.32 input / output · OpenRouter
Context window 128K tokens in · 4.1K out
Since the last one 72 days after Llama 3.2 3B Instruct Capability profile Where it lands among 91 models released in the past year. 50 is the average model.
Reading & documents
No results yet
Science
27 Better than 0% · 1 test
Visual reasoning
No results yet
Computer use
No results yet
Where these scores come from Each capability combines the published benchmarks listed here, rescaled so 50 is the reference average and 15 points is one standard deviation. A hollow marker rests on a single benchmark.
Price against capability The shaded corner holds anything better and cheaper. Prices blend input and output 3 to 1.
2 models from the past year score higher on Science for less: DeepSeek V4 Flash 0731 , Qwen3.7 Flash .
$0.1 $0.3 $1 $3 $10 $30 40 60 80 100 GPT-6.1 Terra: 69, $2 per 1M DeepSeek V4.1 Flash 1017: 69, $0.52 per 1M Muse Spark Open: 66, $2 per 1M Claude Haiku 5.5: 65, $0.2 per 1M Mistral Large 4: 65, $2.06 per 1M Gemini 4 Argon: 71, $4 per 1M GPT-6.1 Sol: 70, $4 per 1M Claude Sonnet 5.5: 76, $4 per 1M GPT-6 Sol: 69, $4 per 1M GPT-6 Luna: 67, $0.2 per 1M Claude Opus 5.5: 108, $8 per 1M Grok 4.7: 61, $3 per 1M DeepSeek V4.1 Flash: 62, $0.21 per 1M GPT-6 Astra: 101, $20 per 1M Gemini 3.8 Flash: 70, $1.5 per 1M Claude Fable 5.1: 90, $20 per 1M GLM-5.3-Flash: 67, $0.24 per 1M GLM-5.3: 60, $2.15 per 1M Gemini 3.7 Flash: 54, $1.5 per 1M DeepSeek V4 Pro 0813: 66, $0.99 per 1M Qwen3.8 2.4T A95B: 70, $3 per 1M Grok 4.6: 58, $3 per 1M Muse Spark 1.2: 60, $2 per 1M Qwen3.8 Max: 68, $3 per 1M DeepSeek V4 Flash 0731: 64, $0.15 per 1M Claude Opus 5: 79, $10 per 1M Gemini 3.6 Flash: 63, $1.5 per 1M Gemini 3.5 Flash Lite: 53, $0.85 per 1M Kimi K3: 62, $6 per 1M Qwen3.7 Flash: 62, $0.055 per 1M GPT-5.6 Terra: 61, $4.5 per 1M GPT-5.6 Sol: 73, $8 per 1M GPT-5.6 Luna: 54, $0.45 per 1M Grok 4.5: 59, $3 per 1M Claude Sonnet 5: 66, $4 per 1M GLM-5.2: 67, $2.15 per 1M Kimi K2.7 Code: 60, $1.71 per 1M Claude Fable 5: 71, $20 per 1M Qwen3.7 Plus: 65, $0.7 per 1M Claude Opus 4.8: 63, $10 per 1M Qwen3.7 Max: 67, $3.75 per 1M Gemini 3.5 Flash: 65, $3.38 per 1M GPT-5.5 Instant: 62, $11.2 per 1M DeepSeek V4 Pro: 51, $1.19 per 1M DeepSeek V4 Flash: 49, $0.18 per 1M GPT-5.5 Pro: 69, $67.5 per 1M GPT-5.5: 73, $11.2 per 1M Qwen3.6 27B: 64, $1.35 per 1M Kimi K2.6: 62, $1.71 per 1M Grok 4.3: 52, $1.56 per 1M Claude Opus 4.7: 70, $10 per 1M Qwen3.6 35B-A3B: 52, $0.84 per 1M GLM-5.1: 62, $2.15 per 1M Gemma 4 31B IT: 57, $0.25 per 1M GPT-5.4 nano: 56, $0.46 per 1M GPT-5.4 mini: 62, $1.69 per 1M Nemotron 3 Super 120B A12B: 44, $0.82 per 1M GPT-5.4 Pro: 75, $67.5 per 1M GPT-5.4: 71, $5.62 per 1M Gemini 3.1 Flash Lite Preview: 54, $0.56 per 1M Gemini 3.1 Flash Lite: 50, $0.56 per 1M Qwen3.5 Flash: 62, $0.18 per 1M Qwen3.5 9B: 59, $0.19 per 1M Qwen3.5 35B-A3B: 62, $0.69 per 1M Qwen3.5 27B: 45, $0.82 per 1M Gemini 3.1 Pro Preview: 73, $4.5 per 1M Claude Sonnet 4.6: 63, $6 per 1M Qwen3.5 Plus: 63, $0.9 per 1M Qwen3.5 397B-A17B: 64, $1.35 per 1M GLM-5: 58, $1.55 per 1M GPT-5.3 Codex: 72, $4.81 per 1M Claude Opus 4.6: 69, $10 per 1M Qwen3 Coder Next: 42, $0.29 per 1M Kimi K2.5: 53, $1 per 1M GLM-4.7: 62, $1 per 1M Gemini 3 Flash Preview: 63, $1.12 per 1M GPT-5.2: 66, $4.81 per 1M DeepSeek V3.2: 62, $0.32 per 1M Claude Opus 4.5: 60, $10 per 1M GPT-5.1: 61, $3.44 per 1M Kimi K2 Thinking: 47, $1.07 per 1M Llama-3.3-70B-Instruct: 27, $0.16 per 1M Llama-3.3-70B-Instruct Price per 1M tokens → ↑ Science $0.1 $1 $10 40 60 80 100 GPT-6.1 Terra: 69, $2 per 1M DeepSeek V4.1 Flash 1017: 69, $0.52 per 1M Muse Spark Open: 66, $2 per 1M Claude Haiku 5.5: 65, $0.2 per 1M Mistral Large 4: 65, $2.06 per 1M Gemini 4 Argon: 71, $4 per 1M GPT-6.1 Sol: 70, $4 per 1M Claude Sonnet 5.5: 76, $4 per 1M GPT-6 Sol: 69, $4 per 1M GPT-6 Luna: 67, $0.2 per 1M Claude Opus 5.5: 108, $8 per 1M Grok 4.7: 61, $3 per 1M DeepSeek V4.1 Flash: 62, $0.21 per 1M GPT-6 Astra: 101, $20 per 1M Gemini 3.8 Flash: 70, $1.5 per 1M Claude Fable 5.1: 90, $20 per 1M GLM-5.3-Flash: 67, $0.24 per 1M GLM-5.3: 60, $2.15 per 1M Gemini 3.7 Flash: 54, $1.5 per 1M DeepSeek V4 Pro 0813: 66, $0.99 per 1M Qwen3.8 2.4T A95B: 70, $3 per 1M Grok 4.6: 58, $3 per 1M Muse Spark 1.2: 60, $2 per 1M Qwen3.8 Max: 68, $3 per 1M DeepSeek V4 Flash 0731: 64, $0.15 per 1M Claude Opus 5: 79, $10 per 1M Gemini 3.6 Flash: 63, $1.5 per 1M Gemini 3.5 Flash Lite: 53, $0.85 per 1M Kimi K3: 62, $6 per 1M Qwen3.7 Flash: 62, $0.055 per 1M GPT-5.6 Terra: 61, $4.5 per 1M GPT-5.6 Sol: 73, $8 per 1M GPT-5.6 Luna: 54, $0.45 per 1M Grok 4.5: 59, $3 per 1M Claude Sonnet 5: 66, $4 per 1M GLM-5.2: 67, $2.15 per 1M Kimi K2.7 Code: 60, $1.71 per 1M Claude Fable 5: 71, $20 per 1M Qwen3.7 Plus: 65, $0.7 per 1M Claude Opus 4.8: 63, $10 per 1M Qwen3.7 Max: 67, $3.75 per 1M Gemini 3.5 Flash: 65, $3.38 per 1M GPT-5.5 Instant: 62, $11.2 per 1M DeepSeek V4 Pro: 51, $1.19 per 1M DeepSeek V4 Flash: 49, $0.18 per 1M GPT-5.5 Pro: 69, $67.5 per 1M GPT-5.5: 73, $11.2 per 1M Qwen3.6 27B: 64, $1.35 per 1M Kimi K2.6: 62, $1.71 per 1M Grok 4.3: 52, $1.56 per 1M Claude Opus 4.7: 70, $10 per 1M Qwen3.6 35B-A3B: 52, $0.84 per 1M GLM-5.1: 62, $2.15 per 1M Gemma 4 31B IT: 57, $0.25 per 1M GPT-5.4 nano: 56, $0.46 per 1M GPT-5.4 mini: 62, $1.69 per 1M Nemotron 3 Super 120B A12B: 44, $0.82 per 1M GPT-5.4 Pro: 75, $67.5 per 1M GPT-5.4: 71, $5.62 per 1M Gemini 3.1 Flash Lite Preview: 54, $0.56 per 1M Gemini 3.1 Flash Lite: 50, $0.56 per 1M Qwen3.5 Flash: 62, $0.18 per 1M Qwen3.5 9B: 59, $0.19 per 1M Qwen3.5 35B-A3B: 62, $0.69 per 1M Qwen3.5 27B: 45, $0.82 per 1M Gemini 3.1 Pro Preview: 73, $4.5 per 1M Claude Sonnet 4.6: 63, $6 per 1M Qwen3.5 Plus: 63, $0.9 per 1M Qwen3.5 397B-A17B: 64, $1.35 per 1M GLM-5: 58, $1.55 per 1M GPT-5.3 Codex: 72, $4.81 per 1M Claude Opus 4.6: 69, $10 per 1M Qwen3 Coder Next: 42, $0.29 per 1M Kimi K2.5: 53, $1 per 1M GLM-4.7: 62, $1 per 1M Gemini 3 Flash Preview: 63, $1.12 per 1M GPT-5.2: 66, $4.81 per 1M DeepSeek V3.2: 62, $0.32 per 1M Claude Opus 4.5: 60, $10 per 1M GPT-5.1: 61, $3.44 per 1M Kimi K2 Thinking: 47, $1.07 per 1M Llama-3.3-70B-Instruct: 27, $0.16 per 1M Llama-3.3-70B-Instruct Price per 1M tokens → ↑ Science OpenRouter $0.1 / $0.32
Since Llama 3.2 3B Instruct 5 releases in the Llama-3.3-70B-Instruct line.
Released 72 days later
Price in / out $0.1 / $0.32 +29%
Context 128K was 131.1K
Meta's past year 8 release days in the past year, about one every 46 days. The next is expected around Dec 15, 2026 . Forecasts ↗
Nov Dec Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Specifications RELEASE DATE Dec 6, 2024
WEIGHTS Open weights
CONTEXT WINDOW 128K
MAX OUTPUT 4.1K
REASONING Not listed
TOOL CALLING Not listed
INPUTS text
OUTPUTS text
Models.dev record ↗
Specifications and prices from Models.dev; availability and limits may differ by provider. Benchmarks from Epoch AI (CC BY 4.0), LiveBench and the other sources named above.
