Skip to main content

MetaLanguage model

Llama 3.1 8B Instruct

Llama 3.1 8B Instruct has no overall index value yet because there are too few independent benchmark results.

PopularityRank 157 of 785

46out of 100-1 vs. last month

based on search, usage and attention

Summary

Llama 3.1 8B Instruct is the version of Llama 3.1 8B tuned for multilingual dialogue. After pretraining, Meta used supervised fine-tuning and reinforcement learning from human feedback.

The larger Llama 3.1 70B Instruct shares its 128K context window. Meta documents tool calls through chat templates and a December 2023 knowledge cutoff. The weights use the Llama 3.1 Community License.

See for yourself: one prompt, one attempt

We show the latest attempt without fixes; it does not count toward the overall index.

Llama 3.1 8B Instruct: Pelican on a bicycle
A community classic for free-form SVG drawing.

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Oct 1, 2026 · 28 s · $0.0000

Test details
Run date
Oct 1, 2026
API provider
openrouter
Requested API model
meta-llama/llama-3.1-8b-instruct
Reasoning setting
not specified
Token limit including reasoning
8,192 tokens
Output tokens
529 tokens
Cost of the attempt
$0.0000
Response time
27.8 s
API model reported in response
meta-llama/llama-3.1-8b-instruct
Pinned endpoint
DeepInfra | meta-llama/llama-3.1-8b-instruct
Temperature
Provider default
Test protocol
community-visual-tests-v1
Technical checks
Passed, not a quality rating

Visually reviewed. This individual test does not establish overall model quality.

Task origin: Simon Willison
SVG short test

SVG short test

One SVG task with the same fixed settings for every model, without hidden thinking.

Llama 3.1 8B Instruct

No independent test. This API route aliases llama-3.1-8b.

All benchmark results

Every published result with its source, test setup and comparison models.

Show all 169 published benchmark results

Page 1 of 15

Measurements without matching peer values

These measurements have no matching peer values under the same test conditions. Their original values and sources remain available here.

MMLU instruct (Show measurement, test conditions, and source)
Source value
69.4
Score
69.4
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table; 5-shot
Source Meta Llama 3.1 model card
MMLU CoT (Show measurement, test conditions, and source)
Source value
73
Score
73
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table; CoT
Source Meta Llama 3.1 model card
MMLU Pro instruct (Show measurement, test conditions, and source)
Source value
48.3
Score
48.3
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
IFEval (Show measurement, test conditions, and source)
Source value
80.4
Score
80.4
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
ARC instruct (Show measurement, test conditions, and source)
Source value
83.4
Score
83.4
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
GPQA (Show measurement, test conditions, and source)
Source value
30.4
Score
30.4
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
HumanEval (Show measurement, test conditions, and source)
Source value
72.6
Score
72.6
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
MBPP (Show measurement, test conditions, and source)
Source value
72.8
Score
72.8
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card
MultiE HumanEval (Show measurement, test conditions, and source)
Source value
50.8
Score
50.8
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct multilingual code table
Source Meta Llama 3.1 model card
MultiE MBPP (Show measurement, test conditions, and source)
Source value
52.4
Score
52.4
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct multilingual code table
Source Meta Llama 3.1 model card
GSM8K (Show measurement, test conditions, and source)
Source value
84.5
Score
84.5
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table; CoT
Source Meta Llama 3.1 model card
MATH (Show measurement, test conditions, and source)
Source value
51.9
Score
51.9
Metric
percent
Unit
%
Direction
Higher is better
Source type
vendor-reported
Evaluator
Meta Llama 3.1 model card
Status
active
Retrieved at
2026-09-08
Methodology
Vendor-reported result. Instruct table
Source Meta Llama 3.1 model card

Page 1 of 15

Popularity compared

How often the model is searched for, used and mentioned, compared with every model of its category.

Popularity index

The 100 most popular models of this category, 0 to 100, higher is more popular

20 of 101 models

GPT-6 Astra83
Claude Haiku 5.577
GPT-6 Sol73
GPT-573
DeepSeek V4.1 Flash 101769
Claude Opus 5.568
Qwen 3.8 27B68
Grok 4.767
Kimi K367
Qwen 3.8 Max 090266
Claude Sonnet 564
Qwen3.5 4B64
DeepSeek-V4-Flash64
Claude Fable 5.164
GPT-5.6 Terra64
MiniCPM-V 4.7 35B-A3B63
GPT-6.1 Sol63
Claude Opus 563
Qwen 3.8 Flash62
Llama 3.1 8B Instruct46

As of November 2026 · Source: gradually.ai popularity index

Over time

Weekly values from October 18, 2026 to November 7, 2026

Llama 3.1 8B InstructMiMo V2.6 Distill Qwen 9B 47Gemma 4 E4B IT 46Sage-A1.5 46Laguna XS.2 46

Method changes in the timeline, a jump at these points is no market event:

  1. 1November 1, 2026: Value only with search or usage data, 30-day attention

What is gaining popularity right now, across models, vendors, harnesses and AI subscriptions, is collected in All AI trends.

Similar models

More models of the same kind as Llama 3.1 8B Instruct, chosen by model line, developer, size, price and release date.

All specifications

Architecture

Parameters
8 billion
Source: Parameters
Knowledge cutoff
December 2023
Source: Knowledge cutoff

Input and output

Context window
131,072 tokens
Source: Context window
Inputs
Text
Source: Inputs
Outputs
Text
Source: Outputs
Tool calling
Yes
Source: Tool calling

Watermarking and provenance

AI watermarking refers to invisible signals embedded in the output. Visible marks and provenance metadata are listed separately. Unknown means there is insufficient evidence for that particular field.

AI watermarking
Unknown
no source
Watermarking technique
Not known
no source

Licence and access

Weights
Open weights
Source: Weights
Access
API
Source: Access
License
Llama 3.1 Community
Source: License

Prices

Not included in any subscription in Germany.

All plans in the AI subscription comparison

Frequently asked questions about Llama 3.1 8B Instruct

What is the knowledge cutoff of Llama 3.1 8B Instruct?

December 2023, according to Meta. That is 7 months before release.

Stated by Meta

How large is the context window of Llama 3.1 8B Instruct?

131,072 tokens. That is how much text Llama 3.1 8B Instruct reads at once, about 200 A4 pages.

Page count as a rule of thumb from Artificial Analysis

Context windowTokens the model reads at once · higher is better
GLM-5.1200K
DeepSeek-V3 0324164K
Llama 3.1 8B InstructRank 94 of 103131K
GLM-4.5131K
GPT OSS 120B131K
DeepSeek-V3.2128K
DeepSeek-R1128K

Can I run Llama 3.1 8B Instruct locally?

Yes, the weights are open.

Stated by Meta

Open weights
yes
License
Llama 3.1 Community
Parameters
8B
active per token
n/a

Sources and data date

Show all 40 sources

Benchmark sources at a glance

Which independent evaluators tested this model. This is coverage, not a quality ranking.

EvaluatorTest seriesComposite indicesRecorded active measurements
LMArena001
Berkeley Function Calling Leaderboard001

Without vendor claims and historical or disputed results; those stay labelled in the single values. 2 measurements are not yet assigned to a test series.

TypeEvidence and data date
Input

Documented modalities.

Source: Hugging Face · Hugging Face (retrieved October 15, 2026; October 17, 2026) · Hugging Face, Llama 3.1 8B Instruct · Editorial description reviewed October 4, 2026
Output

Documented modalities.

Source: Hugging Face · Hugging Face (retrieved October 15, 2026; October 17, 2026) · Hugging Face, Llama 3.1 8B Instruct · Editorial description reviewed October 4, 2026
Research statusResearch date September 8, 2026. 34 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not published

external-benchmark-row-level-uncertainty

Note

CRADLE, SHINES, and MINDBENCH do not provide confidence intervals for these exact rows. SHINES labels its parenthesized dispersion as variance, which is not converted into a confidence interval. Effective sample counts are assigned only for MINDBENCH, whose Table 5 explicitly reports 3,000 and 2,999.

Not published

per-row evaluation N and confidence interval

Note

The paper describes dataset sizes but reports no model-row sample size or confidence interval.

Unresolved

row-specific prompt, route, and checkpoint revision

Note

Study-level protocol is described, but no prompt variant, serving route, or checkpoint revision is bound to this row.

Not published

evaluated N

Original wording

The leaderboard reports preference-vote totals but no fixed task-case N; Votes are not copied to sampleCount.

Unresolved

confidence interval

Original wording

The Score cell displays ±, but the official policy does not define that notation as a confidence interval; no CI field is populated.

Unresolved

score direction

Original wording

The captured source does not explicitly state higher-is-better or lower-is-better; higherIsBetter remains null.

Not published

per-model prompts and generation settings

Original wording

No exact prompt, decoding settings, or serving configuration is published in the captured row.

Additional sourceMeta Llama 3.1 model card (retrieved September 8, 2026)
Additional sourceCRADLE Bench: A Clinician-Annotated Benchmark for Multi-Faceted Mental Health Crisis and Safety Risk Detection (retrieved October 17, 2026)
Additional sourceACL Anthology 2025.acl-long.1330, SHINES paper (retrieved October 17, 2026)
Additional sourceUniversity of Oulu thesis, MINDBENCH: A Validated Benchmark Dataset for Evaluating LLM Performance in Mental Health Disorder Classification and Counseling Dialogue (retrieved October 17, 2026)
Additional sourceConvCodeWorld: Benchmarking Conversational Code Generation in Reproducible Feedback Environments (retrieved October 17, 2026)
Additional sourceACTIONREASONINGBENCH: Reasoning about Actions with and without Ramification Constraints (retrieved October 17, 2026)
Additional sourceFrom Noise to Signal: Improving Security Log Anomaly Detection Using LLMs with Endpoint-Specific Logs (arXiv:2608.19938v1) (retrieved October 17, 2026)
Additional sourceSciTaRC, Wang et al. (retrieved October 17, 2026)
Additional sourceWhich Quantization Should I Use? A Unified Evaluation of llama.cpp Quantization on Llama-3.1-8B-Instruct (retrieved October 18, 2026)
Additional source6G-Bench: An Open Benchmark for Semantic Communication and Network-Level Reasoning with Foundation Models in AI-Native 6G Networks (retrieved October 18, 2026)
Additional sourceACL Anthology, Findings of EMNLP 2025, From KMMLU-Redux to Pro (retrieved October 18, 2026)
Additional sourceArena leaderboard (retrieved October 18, 2026)
Additional sourceSocioBench authors, Table 3 (retrieved October 19, 2026)
Additional sourceLongBioBench maintainers, results table (retrieved October 19, 2026)
Additional sourceOpenCompass GPassK maintainers, Main Results table (retrieved October 19, 2026)
Additional sourceProbing Preference Representations (arXiv:2511.12464v1) (retrieved October 22, 2026)
Additional sourceKhalid Yusuf Dahir, SomaliBench v0 (retrieved October 22, 2026)
Additional sourceArdana JevBench (retrieved October 22, 2026)
Additional sourceCACHENOTES authors, EACL 2026 (retrieved October 23, 2026)
Additional sourceBigCodeBench official leaderboard (retrieved October 28, 2026)12
Additional sourceBerkeley Function Calling Leaderboard V4 official overall CSV (retrieved October 28, 2026)
Additional sourceModel metadata (retrieved September 8, 2026)
Additional sourceModel metadata (retrieved October 17, 2026)
Additional sourceModel metadata (retrieved October 18, 2026)
Additional sourceModel metadata (retrieved October 23, 2026)
Additional sourceModel metadata (retrieved October 28, 2026)1234567
Additional sourceOpenRouter: Popularity index (derived), CC BY 4.0
Additional sourceVercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional sourceOllama: cumulative downloads for the popularity index
Additional sourceReplicate: cumulative runs for the popularity index
Additional sourceGoogle Trending Now: limited feed selection for Germany and the US
Additional sourceCreative Commons Attribution 4.0