Meta model product
Llama 3.1 8B
1 8B Instruct is a multilingual large language model optimized for dialogue use cases.
Updated Aug 12, 2026. Default version: Llama 3.1 8B Instruct
LLMBoard score0.0Llama 3.1 8B Instruct
Coverage100%16 benchmark families
Context window131.1KTokens
Official input priceN/AOfficial price unavailable
Llama 3.1 8B Specifications
Technical details for the model's default version.
- Version
- Llama 3.1 8B Instruct
- Released
- Jul 23, 2024
- Knowledge cutoff
- Dec 31, 2023
- Parameters
- 8B
- Context window
- 131.1K
- Max output
- 131.1K
- Inputs
- text
- Outputs
- text
- Open weights
- No
- License
- Llama 3.1 Community License
Llama 3.1 8B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Llama 3.1 8B Benchmark Results
Benchmark scores for Llama 3.1 8B Instruct.
| MBPP EvalPlus (base) | 72.8% | 01 | 1 | 100.0% | C | |
| GSM-8K (CoT) | 84.5% | 02 | 2 | 0.0% | C | |
| API-Bank | 82.6% | 03 | 3 | 0.0% | C | |
| BFCL | 76.1% | 03 | 11 | 80.0% | C | |
| Gorilla Benchmark API Bench | 8.2% | 03 | 3 | 0.0% | C | |
| MMLU (CoT) | 73.0% | 03 | 3 | 0.0% | C | |
| Multilingual MGSM (CoT) | 68.9% | 03 | 3 | 0.0% | C | |
| Multipl-E HumanEval | 50.8% | 03 | 3 | 0.0% | C | |
| Multipl-E MBPP | 52.4% | 03 | 3 | 0.0% | C | |
| Nexus | 38.5% | 03 | 4 | 33.3% | C | |
| MATH (CoT) | 51.9% | 06 | 6 | 0.0% | C | |
| ARC-C | 83.4% | 16 | 34 | 54.5% | C | |
| DROP | 59.5% | 24 | 30 | 20.7% | C | |
| IFEval | 80.4% | 52 | 65 | 20.3% | C | |
| HumanEval | 72.6% | 54 | 66 | 18.5% | C | |
| MMLU | 69.4% | 85 | 100 | 15.2% | C | |
| MMLU-Pro | 48.3% | 116 | 129 | 10.2% | C | |
| GPQA | 30.4% | 224 | 234 | 4.3% | C |
Llama 3.1 8B Arena Results
Preference and agent-evaluation results for the default version.
| text | overall | 297 | 1186.7 | 49,605 | N/A | |
| text style control | overall | 313 | 1211.3 | 49,605 | N/A |
Llama 3.1 8B Runtime Performance
Provider-specific output speed and catalog latency for Llama 3.1 8B Instruct. Runtime does not affect the capability score.
| Cerebras | 2,047 tok/s | 0.2 s | 131.1K | 131.1K | |
| Sambanova | 1,050 tok/s | 0.5 s | 131.1K | 131.1K | |
| Groq | 750 tok/s | 0.5 s | 131.1K | 131.1K | |
| Fireworks | 292 tok/s | 0.5 s | 131.1K | 131.1K | |
| Hyperbolic | 200 tok/s | 0.5 s | 131.1K | 131.1K | |
| Together | 194 tok/s | 0.5 s | 131.1K | 131.1K | |
| DeepInfra | 118 tok/s | 0.5 s | 131.1K | 131.1K | |
| Bedrock | 100 tok/s | 0.5 s | 131.1K | 131.1K | |
| Lambda | 42 tok/s | 0.5 s | 131.1K | 131.1K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Llama 3.1 8B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.02 input, $0.05 output per 1M via Helicone
- Tracked offerings
- 13
| Nvidia | meta/llama-3.1-8b-instruct | global | N/A | N/A | 16K | |
| Helicone | llama-3.1-8b-instruct | global | $0.02 | $0.05 | 16.4K | |
| Abacus | meta-llama/Meta-Llama-3.1-8B-Instruct | global | $0.02 | $0.05 | 128K | |
| NovitaAI | meta-llama/llama-3.1-8b-instruct | global | $0.02 | $0.05 | 16.4K | |
| Inference | meta/llama-3.1-8b-instruct | global | $0.025 | $0.025 | 16K | |
| OpenRouter | meta-llama/llama-3.1-8b-instruct | global | $0.05 | $0.08 | 131.1K | |
| NanoGPT | meta-llama/llama-3.1-8b-instruct | global | $0.0544 | $0.085 | 131.1K | |
| Neon | meta-llama-3-1-8b-instruct | global | $0.15 | $0.45 | 131.1K | |
| Cortecs | llama-3.1-8b-instruct | global | $0.167 | $0.167 | 128K | |
| Pioneer | meta-llama/Llama-3.1-8B-Instruct | global | $0.20 | $0.20 | 131.1K | |
| Amazon Bedrock | meta.llama3-1-8b-instruct-v1:0 | global | $0.22 | $0.22 | 128K | |
| Vercel AI Gateway | meta/llama-3.1-8b | global | $0.22 | $0.22 | 128K | |
| Cloudflare AI Gateway | workers-ai/@cf/meta/llama-3.1-8b-instruct | global | $0.28 | $0.83 | 128K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Llama 3.1 8B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Llama 3.1 8B Instruct | 0.0 | 8B | 131.1K | 131.1K | No | Llama 3.1 Community License |
What is Llama 3.1 8B?
Key information about Llama 3.1 8B and its available data.
1 8B Instruct is a multilingual large language model optimized for dialogue use cases. It features a 128K context length, state-of-the-art tool use, and strong reasoning capabilities.
Data as of 2026-08-11.
Llama 3.1 8B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Llama 3.1 8B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
ME
Llama 3.2 3B
Meta
ME
Llama 3.2 11B
Meta
ME
Llama 3.2 90B
Meta
ME
Llama 3.1 70B
Meta
ME
Llama 4 Scout
Meta
ME
Llama 3.3 70B
Meta
FAQ
Common questions about Llama 3.1 8B.
When was Llama 3.1 8B released?
Llama 3.1 8B's default version was released on Jul 23, 2024.
How much does Llama 3.1 8B cost?
No official standard PAYG price is currently available for Llama 3.1 8B. The lowest tracked third-party offer starts at $0.02 input and $0.05 output via Helicone.
Who created Llama 3.1 8B?
Llama 3.1 8B was created by Meta.
What is the context window for Llama 3.1 8B?
The default version has a 131.1K token context window.
Is Llama 3.1 8B open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer Llama 3.1 8B?
13 provider offerings are linked to the default version.
What models should I compare Llama 3.1 8B with?
Nearby ranked alternatives include Command R+, Gemma 3 1B, Gemini 1.0 Pro.