Automated page. Written and translated automatically by AI and checked against the sources by software; no human editor. SI-IQ is an index built from public benchmark results, not an IQ test. How it works

How smart is Llama 3.3 70B?

66 SI-IQ · 90% interval 61–71
Rank 80 of 80 ranked models · Plausible rank range 77–80

Llama 3.3 70B is a model from Meta AI, released on Dec 6, 2024. Its SI-IQ is 66 (90% interval 61–71), estimated from 5 benchmark results by 3 independent evaluators. Its strongest area is Agents and its weakest is Math.

Facts

Developer: Meta AI
Released: Dec 6, 2024
Access: Open weights (restricted licence)
Benchmarks used: 5 results from 3 evaluators

Ability by area (SI-IQ, 90% interval)

Every benchmark result we used

Each score is one value exactly as the source published it: for Epoch AI the best of its run settings, for OpenRouter the median hosting provider (the developer's own hosting when it ran the test). “Implied SI-IQ” is the ability that this single result points to; the overall SI-IQ is their precision-weighted average.

Benchmark	Area	Score	Implied SI-IQ	Evaluated by	As listed by the source	Retrieved	Source
MATH Level 5	Math	41.6%	65	Epoch AI	`Llama-3.3-70B-Instruct`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
OTIS Mock AIME 2024–2025	Math	5.1%	63	Epoch AI	`Llama-3.3-70B-Instruct`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
GPQA Diamond	Knowledge	47.4%	70	Epoch AI	`Llama-3.3-70B-Instruct`	Oct 18, 2026	Epoch AI Benchmarking Hub · data file
GPQA Diamond (OpenRouter run)	Knowledge	47.5%	70	OpenRouter	`meta-llama/llama-3.3-70b-instruct` Parasail · median over providers (12 of 12 rows), lower median	Oct 18, 2026	OpenRouter benchmark datasets · data file
Berkeley Function Calling Leaderboard	Agents	31.9%	73	UC Berkeley (Gorilla)	`Llama-3.3-70B-Instruct (FC)`	Oct 18, 2026	Berkeley Function Calling Leaderboard · data file
τ²-bench Airline (OpenRouter run)	Agents	31.3%	61	OpenRouter	`meta-llama/llama-3.3-70b-instruct` PrimeIntellect · median over providers (6 of 6 rows), lower median	Oct 18, 2026	OpenRouter benchmark datasets · data file

Human preference (Arena)

On the Arena leaderboard, where people vote between two anonymous answers, Llama 3.3 70B has a rating of 1,318 (1,315–1,322) from 54,497 votes (published Oct 15, 2026). Human preference is a different measure from benchmark ability and is not part of SI-IQ. Arena (Hugging Face)

Models with similar SI-IQ

Model	SI-IQ
GPT-4o miniOpenAI	68	Llama 3.3 70B vs GPT-4o mini
GPT-4o (Nov 2024)OpenAI	69	Llama 3.3 70B vs GPT-4o (Nov 2024)
GPT-4o (Aug 2024)OpenAI	73	Llama 3.3 70B vs GPT-4o (Aug 2024)
GPT-4.1 nanoOpenAI	76	Llama 3.3 70B vs GPT-4.1 nano
Gemma 3 27BGoogle DeepMind	77	Llama 3.3 70B vs Gemma 3 27B

Computed with SI-IQ method version 1.0.0. Read the full methodology

Data sources: Arena leaderboard dataset on Hugging Face (CC BY 4.0); Berkeley Function Calling Leaderboard, Gorilla project (Apache-2.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..