MetaLanguage model

Llama-3.3-70B-Instruct

Llama-3.3-70B-Instruct has no overall index value yet because there are too few independent benchmark results.

Intelligencenot ranked yet

n/aOverall indexhigher is better

SpeedRank 62 of 101

75Tokens per secondhigher is better

Priceno API price

n/aper 1M tokenslower is better

ContextRank 98 of 103

128Ktokens at once≈ 190 A4 pages

PopularityRank 98 of 785

50out of 100-1 vs. last month

based on search, usage and attention

Summary

At 75 tokens per second it writes slower than a typical model (87).

Llama 3.3 70B Instruct is the open Instruct checkpoint of the Llama 3.3 70B generation. The model card lists tool calling and the Llama 3.3 Community License. Llama 3.3 70B Instruct is documented as a self-hostable weight checkpoint.

Llama 3.3 70B is the parallel directory entry for the same generation. The card gives a 128,000-token context window and a December 2023 knowledge cutoff.

See for yourself: one prompt, one attempt

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

Pelican on a bicycle

Prompt

Generate an SVG of a pelican riding a bicycle

Oct 1, 2026 · 20 s · $0.0002

Test details

Run date: Oct 1, 2026
API provider: openrouter
Requested API model: meta-llama/llama-3.3-70b-instruct
Reasoning setting: not specified
Token limit including reasoning: 8,192 tokens
Output tokens: 663 tokens
Cost of the attempt: $0.0002
Response time: 20.4 s
API model reported in response: meta-llama/llama-3.3-70b-instruct
Pinned endpoint: DeepInfra | meta-llama/llama-3.3-70b-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

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

Llama-3.3-70B-Instruct

No independent test. This API route aliases llama-3.3-70b.

All benchmark results

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

Show all 80 published benchmark results

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.

CritPt v2: Overall (Show measurement, test conditions, and source)

Source value: 0
Score: 0
Metric: pass@1 accuracy
Unit: %
Benchmark version: v2
Category: reasoning
Task: overall
Task label: Overall
Direction: Higher is better
Source type: independent-evaluator
Evaluator: Artificial Analysis
Status: disputed
Retrieved at: 2026-11-02
Methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Source model ID: artificial-analysis/976cc8ad-7904-4056-83c5-960181f47d5f/llama-3-3-instruct-70b/llama-3-3-instruct-70b. Source model name: Llama-3.3-70B-Instruct. Provider reported by source: Meta. Model snapshot: Llama 3.3 Instruct 70B. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: non-reasoning. Reasoning effort: none. Score display: 0.0%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-critpt-v2:v2:standalone:non-reasoning:effort=unspecified:suffix=unspecified
Comparison methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. The full CritPt v2 run covers 70 questions with five repeats. The official grading server reports pass@1.. Tool policy: Python functions, symbolic expressions, and numerical answers; official grading server. Attempts: 70 questions, five repeats per question, pass@1. Reasoning: non-reasoning. Reasoning effort: none
Context: This AA source field contains the original CritPt score. The earlier CritPt v2 attribution was incorrect; the value and former label remain as disputed historical evidence.. The source dataset marks this result as disputed. It remains visible as documented evidence, but is not reliable comparative proof.

Source Artificial Analysis

IFBench: Overall (Show measurement, test conditions, and source)

Source value: 47.0748299319728
Score: 47.0748299319728
Metric: IFBench prompt-level loose accuracy (pass@1)
Unit: %
Benchmark version: AA test set: 294 questions
Category: general
Task: overall
Task label: Overall
Direction: Higher is better
Source type: independent-evaluator
Evaluator: Artificial Analysis
Status: superseded
Retrieved at: 2026-11-02
Methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/976cc8ad-7904-4056-83c5-960181f47d5f/llama-3-3-instruct-70b/llama-3-3-instruct-70b. Source model name: Llama-3.3-70B-Instruct. Provider reported by source: Meta. Model snapshot: Llama 3.3 Instruct 70B. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 47.1%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:non-reasoning:effort=unspecified:suffix=unspecified
Comparison methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none
Context: The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.. This result is retained as historical evidence and is not interchangeable with current results.

Source Artificial Analysis

IFBench: Overall (Show measurement, test conditions, and source)

Source value: 47.0748299319728
Score: 47.0748299319728
Metric: IFBench prompt-level loose accuracy (pass@1)
Unit: %
Benchmark version: AA test set: 294 questions
Category: general
Task: overall
Task label: Overall
Direction: Higher is better
Source type: independent-evaluator
Evaluator: Artificial Analysis
Status: active
Retrieved at: 2026-11-07
Methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Source model ID: artificial-analysis/976cc8ad-7904-4056-83c5-960181f47d5f/llama-3-3-instruct-70b/llama-3-3-instruct-70b. Source model name: Llama-3.3-70B-Instruct. Provider reported by source: Meta. Model snapshot: Llama 3.3 Instruct 70B. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none. Score display: 47.1%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-07-b96e3ffc8c5b. Source payload hash: b96e3ffc8c5b052aa56f96d05e89d02b8cc234a75085820fca0ca2a39e498274. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-07. Source cohort: artificial-analysis-ifbench:AA test set: 294 questions:standalone:non-reasoning:effort=unspecified:suffix=unspecified
Comparison methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. Artificial Analysis runs 294 questions from the IFBench test set with five repeats per question. The official evaluator scores prompt-level accuracy in loose mode as pass@1.. Tool policy: Single-turn answer evaluated with the official IFBench source code. Attempts: pass@1, 5 repeats per question. Reasoning: non-reasoning. Reasoning effort: none
Context: The often cited 58 refers to verifiable out-of-domain constraint types, not the number of questions in AA's test run.

Source Artificial Analysis

ITBench-AA (SRE): Overall (Show measurement, test conditions, and source)

Source value: 0.5649717514124291
Score: 0.5649717514124291
Metric: Average precision at full recall
Unit: %
Benchmark version: AA cohort: 59 tasks
Category: agentic
Task: overall
Task label: Overall
Direction: Higher is better
Source type: independent-evaluator
Evaluator: Artificial Analysis
Status: superseded
Retrieved at: 2026-11-02
Methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. 59 Kubernetes incident tasks, 40 public and 19 private, with three repeats per task. A run scores zero if it misses any true root cause; otherwise it receives its precision at full recall. This is the SRE slice of ITBench.. Source model ID: artificial-analysis/976cc8ad-7904-4056-83c5-960181f47d5f/llama-3-3-instruct-70b/llama-3-3-instruct-70b. Source model name: Llama-3.3-70B-Instruct. Provider reported by source: Meta. Model snapshot: Llama 3.3 Instruct 70B. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Harness: Artificial Analysis Stirrup. Agent: Stirrup agent. Tool policy: Shell access to an offline Kubernetes incident snapshot. Attempts: Average precision at full recall, 3 repeats per task. Reasoning: non-reasoning. Reasoning effort: none. Score display: 0.6%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-02-67d46be20a5b. Source payload hash: 67d46be20a5b412db6cfb60ad190441fcaa5513500b8fd47d033168d0e97b6b2. Parser version: artificial-analysis-rsc-models-v3. Imported at: 2026-11-02. Source cohort: artificial-analysis-itbench-aa:AA cohort: 59 tasks:standalone:non-reasoning:effort=unspecified:suffix=unspecified
Comparison methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). AA marks this model variant as non-reasoning and reports 0 reasoning tokens. 59 Kubernetes incident tasks, 40 public and 19 private, with three repeats per task. A run scores zero if it misses any true root cause; otherwise it receives its precision at full recall. This is the SRE slice of ITBench.. Harness: Artificial Analysis Stirrup. Agent: Stirrup agent. Tool policy: Shell access to an offline Kubernetes incident snapshot. Attempts: Average precision at full recall, 3 repeats per task. Reasoning: non-reasoning. Reasoning effort: none
Context: This AA implementation covers Site Reliability Engineering only, not ITBench's other CISO and FinOps domains.. This result is retained as historical evidence and is not interchangeable with current results.

Source Artificial Analysis

ITBench-AA (SRE): Overall (Show measurement, test conditions, and source)

Source value: 0.5649717514124291
Score: 0.5649717514124291
Metric: Average precision at full recall
Unit: %
Benchmark version: AA cohort: 59 tasks
Category: agentic
Task: overall
Task label: Overall
Direction: Higher is better
Source type: independent-evaluator
Evaluator: Artificial Analysis
Status: active
Retrieved at: 2026-11-07
Methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. 59 Kubernetes incident tasks, 40 public and 19 private, with three repeats per task. A run scores zero if it misses any true root cause; otherwise it receives its precision at full recall. This is the SRE slice of ITBench.. Source model ID: artificial-analysis/976cc8ad-7904-4056-83c5-960181f47d5f/llama-3-3-instruct-70b/llama-3-3-instruct-70b. Source model name: Llama-3.3-70B-Instruct. Provider reported by source: Meta. Model snapshot: Llama 3.3 Instruct 70B. Evaluator: Artificial Analysis. Dataset type: artificial-analysis-evaluation. Harness: Artificial Analysis Stirrup. Agent: Stirrup agent. Tool policy: Shell access to an offline Kubernetes incident snapshot. Attempts: Average precision at full recall, 3 repeats per task. Reasoning: non-reasoning. Reasoning effort: none. Score display: 0.6%. Source ID: artificial-analysis. Evidence capture ID: aa-model-page-2026-11-07-b96e3ffc8c5b. Source payload hash: b96e3ffc8c5b052aa56f96d05e89d02b8cc234a75085820fca0ca2a39e498274. Parser version: artificial-analysis-rsc-models-v4. Imported at: 2026-11-07. Source cohort: artificial-analysis-itbench-aa:AA cohort: 59 tasks:standalone:non-reasoning:effort=unspecified:suffix=unspecified
Comparison methodology: Exact AA configuration: Llama 3.3 Instruct 70B (release llama-3-3-instruct-70b, row llama-3-3-instruct-70b, UUID 976cc8ad-7904-4056-83c5-960181f47d5f, reasoning=false, effort=not specified, suffix=not specified). This source capture does not provide an average reasoning-token count. 59 Kubernetes incident tasks, 40 public and 19 private, with three repeats per task. A run scores zero if it misses any true root cause; otherwise it receives its precision at full recall. This is the SRE slice of ITBench.. Harness: Artificial Analysis Stirrup. Agent: Stirrup agent. Tool policy: Shell access to an offline Kubernetes incident snapshot. Attempts: Average precision at full recall, 3 repeats per task. Reasoning: non-reasoning. Reasoning effort: none
Context: This AA implementation covers Site Reliability Engineering only, not ITBench's other CISO and FinOps domains.

Source Artificial Analysis

MMLU CoT (Show measurement, test conditions, and source)

Source value: 86
Score: 86
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table; chain-of-thought setting as labeled

Source Meta Llama 3.3 model card

MMLU Pro (Show measurement, test conditions, and source)

Source value: 68.9
Score: 68.9
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

IFEval (Show measurement, test conditions, and source)

Source value: 92.1
Score: 92.1
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

GPQA Diamond (Show measurement, test conditions, and source)

Source value: 50.5
Score: 50.5
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

HumanEval (Show measurement, test conditions, and source)

Source value: 88.4
Score: 88.4
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

MBPP (Show measurement, test conditions, and source)

Source value: 87.6
Score: 87.6
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

MATH (Show measurement, test conditions, and source)

Source value: 77
Score: 77
Metric: percent
Unit: %
Direction: Higher is better
Source type: vendor-reported
Evaluator: Meta Llama 3.3 model card
Status: active
Retrieved at: 2026-09-08
Methodology: Vendor-reported result. Instruction-tuned table; Card benchmark table

Source Meta Llama 3.3 model card

Page 1 of 7

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 100 models

GPT-6 Astra83

Claude Haiku 5.577

GPT-6 Sol73

GPT-573

DeepSeek V4.1 Flash 101769

Kimi K367

MiniCPM-V 4.7 35B-A3B63

GPT-6.1 Sol63

Claude Opus 563

Qwen 3.8 Flash62

Llama-3.3-70B-Instruct50

DeepSeek V4.1 Flash 1017

MiniCPM-V 4.7 35B-A3B

GPT-6.1 Sol

Claude Opus 5

Qwen 3.8 Flash

Llama-3.3-70B-Instruct

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

Over time

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

Llama-3.3-70B-InstructGemma 4 26B A4B 51Gemini 3.1 Pro Preview 50Flint 50Claude Sonnet 4.5 50

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

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-3.3-70B-Instruct, chosen by model line, developer, size, price and release date.

All specifications

Architecture

Parameters: 70 billion; Source: Parameters
Knowledge cutoff: December 2023; Source: Knowledge cutoff

Input and output

Context window: 128,000 tokens; Source: Context window
Maximum output: 4,096 tokens; Source: Maximum output
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 3.3 Community; Source: License

Prices

Not included in any subscription in Germany.

All plans in the AI subscription comparison

Frequently asked questions about Llama-3.3-70B-Instruct

What is the knowledge cutoff of Llama-3.3-70B-Instruct?

December 2023, according to Meta. That is 12 months before release.

Stated by Meta

DecFebAprJunAugOctDecKnowledge cutoffDecember 2023ReleasedDec 6, 2024

How large is the context window of Llama-3.3-70B-Instruct?

128,000 tokens. That is how much text Llama-3.3-70B-Instruct reads at once, about 190 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

GLM-4.5131K

GPT OSS 120B131K

Llama-3.3-70B-InstructRank 98 of 103128K

DeepSeek-V3.2128K

DeepSeek-R1128K

How fast is Llama-3.3-70B-Instruct?

75 tokens per second. That is slower than a typical model at 87.

Measured by Artificial Analysis on Nov 6, 2026. AA 3.0.0, including time to first token, averaged across available workloads and five days.

SpeedTokens per second · higher is better

GPT-5.579

Mistral Large 378

Llama-3.3-70B-InstructRank 62 of 10175

Gemini 4 Argon74

GPT-6.1 Terra73

Grok 4.672

Grok 4.771

Can I run Llama-3.3-70B-Instruct locally?

Yes, the weights are open.

Stated by Meta

Open weights: yes
License: Llama 3.3 Community
Parameters: 70B
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.

Evaluator	Test series	Recorded active measurements
Artificial Analysis	3	3
LMArena	0	1
Berkeley Function Calling Leaderboard	0	1

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.

Type	Evidence and data date
Speed	Median across serving providers, Non-reasoning. AA 3.0.0, including time to first token, averaged across available workloads and five days. Retrieved November 6, 2026. Source: Artificial Analysis · Model metadata (retrieved September 8, 2026)Speed measurement methodology
Input	Documented modalities. Retrieved October 27, 2026. Source: models.dev catalog · models.dev catalog (retrieved October 27, 2026) · Meta model catalog snapshot · Editorial description reviewed October 2, 2026
Output	Documented modalities. Retrieved October 27, 2026. Source: models.dev catalog · models.dev catalog (retrieved October 27, 2026) · Meta model catalog snapshot · Editorial description reviewed October 2, 2026
Research status	Research date September 8, 2026. 28 source URLs checked. This documents the inspected sources, not an exhaustive inventory of every publication.
Not published	effective-sample-count-and-confidence-interval Note CRADLE and SciTaRC do not provide an effective model-row sample count or confidence interval. SciTaRC's 245 manually annotated judge-agreement questions are not the model-scoring denominator and were not imported as N.
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 source	Hugging Face (retrieved October 15, 2026) · Hugging Face, Llama 3.3 70B Instruct · Editorial description reviewed October 2, 2026
Additional source	Artificial Analysis (retrieved November 2, 2026)
Additional source	Artificial Analysis (retrieved October 29, 2026; November 2, 2026; November 7, 2026)
Additional source	Artificial Analysis (retrieved November 2, 2026; November 7, 2026)1 2
Additional source	Meta Llama 3.3 model card (retrieved September 8, 2026)
Additional source	CRADLE Bench: A Clinician-Annotated Benchmark for Multi-Faceted Mental Health Crisis and Safety Risk Detection (retrieved October 17, 2026)
Additional source	SciTaRC: A Plan-Annotated Scientific Tabular QA Benchmark for Language Reasoning and Complex Computation (retrieved October 17, 2026)
Additional source	Robustness as an Emergent Property of Task Performance (retrieved October 17, 2026)
Additional source	How Good Are LLMs at Processing Tool Outputs? (retrieved October 17, 2026)
Additional source	Assessing Large Language Models for Medical QA: Zero-Shot and LLM-as-a-Judge Evaluation (arXiv:2602.14564v1) (retrieved October 17, 2026)
Additional source	CausalVerify, Zhang et al. (retrieved October 17, 2026)
Additional source	Could language models win the International Linguistics Olympiad? (retrieved October 18, 2026)
Additional source	ACL Anthology, Findings of EMNLP 2025, From KMMLU-Redux to Pro (retrieved October 18, 2026)
Additional source	Arena leaderboard (retrieved October 18, 2026)
Additional source	DynaCode paper, Findings of ACL 2025, ACL Anthology ID 2025.findings-acl.1133 (retrieved October 19, 2026)
Additional source	Are LLMs Ready for Computer Science Education? A Cross-Domain, Cross-Lingual and Cognitive-Level Evaluation Using Professional Certification Exams (arXiv:2604.06898v1) (retrieved October 19, 2026)
Additional source	Oracle Cloud Infrastructure (retrieved October 20, 2026)
Additional source	BigCodeBench official leaderboard (retrieved October 28, 2026)1 2
Additional source	Berkeley Function Calling Leaderboard V4 official overall CSV (retrieved October 28, 2026)
Additional source	Model metadata (retrieved October 17, 2026)1 2
Additional source	Model metadata (retrieved October 18, 2026)
Additional source	Model metadata (retrieved October 28, 2026)1 2 3 4 5 6 7
Additional source	OpenRouter: Popularity index (derived), CC BY 4.0
Additional source	Vercel AI Gateway: Popularity index (derived), CC BY 4.0
Additional source	Ollama: cumulative downloads for the popularity index
Additional source	Replicate: cumulative runs for the popularity index
Additional source	Google Trending Now: limited feed selection for Germany and the US
Additional source	Creative Commons Attribution 4.0