Llama 3.1 70B Instruct vs Llama 3.3 70B Instruct
Wins 1 of 6 areas
Facts
Wins 3 of 6 areas
Reasoning · Long documents · Following instructions
Llama 3.3 70B Instruct wins more areas, narrowly.Llama 3.1 70B Instruct is cheaper and better at facts.
Scores updated · 30 tests both models report · How we compare
Where each one wins
Tests won in each of the six areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking01Llama 3.3 70B Instruct1 of 3 tests · 2 ties
- Long documentsFinding answers in very long texts01Llama 3.3 70B Instruct1 of 1 test
- Following instructionsDoing exactly what it is asked01Llama 3.3 70B Instruct1 of 1 test
- FactsGetting facts right instead of making them up10Llama 3.1 70B Instruct1 of 2 tests · 1 tie
- CodingWriting and fixing software00Even0 each · 2 ties
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
Agents, long documents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
Llama 3.1 70B Instruct costs 22% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Llama 3.3 70B Instruct pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+12.7points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+8.9points ahead
- Reasons across sets of long documentsAA-LCR+7.7points ahead
Where Llama 3.1 70B Instruct pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+12points ahead
Every test, side by side
All 30 tests both models report. The winning score is in its model's colour; marks a score checked independently.
ReasoningLlama 3.3 70B Instruct
- GPQA DiamondLlama 3.3 70B Instruct by 8.940.949.8+8.9
- Humanity's Last Examtie4.53.6tie
- CritPttie00tie
FactsLlama 3.1 70B Instruct
- AA-Omniscience · Non-hallucinationLlama 3.1 70B Instruct by 1221.89.8+12
- AA-Omniscience · Accuracytie19.718.9tie
Long documentsLlama 3.3 70B Instruct
- AA-LCRLlama 3.3 70B Instruct by 7.7815.7+7.7
Following instructionsLlama 3.3 70B Instruct
- IFBenchLlama 3.3 70B Instruct by 12.734.447.1+12.7
Other results20 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- τ²-Bench Telecom (AA run)Llama 3.3 70B Instruct by 11.415.226.6+11.4
- AA-OmniscienceLlama 3.1 70B Instruct by 11.1-43.1-54.2+11.1
- DROPLlama 3.3 70B Instruct by 10.679.690.2+10.6
- MATHLlama 3.3 70B Instruct by 96877+9
- MATH (CoT) (Pass@1 with chain-of-thought)Llama 3.3 70B Instruct by 96877+9
- HumanEvalLlama 3.3 70B Instruct by 7.980.588.4+7.9
- IFEvalLlama 3.3 70B Instruct by 4.687.592.1+4.6
- MGSMLlama 3.3 70B Instruct by 4.286.991.1+4.2
- MMLU Pro (CoT) (Pass@1 with chain-of-thought)Llama 3.3 70B Instruct by 2.566.468.9+2.5
- MMLU-ProLlama 3.3 70B Instruct by 2.566.468.9+2.5
- naturalquestions_closedbookLlama 3.1 70B Instruct by 2.145.243.1+2.1
- NarrativeQALlama 3.3 70B Instruct by 1.977.279.1+1.9
- MBPP EvalPlus (base)Llama 3.3 70B Instruct by 1.68687.6+1.6
- AA IntelligenceLlama 3.3 70B Instruct by 1.16.67.7+1.1
- Artificial Analysis Coding IndexLlama 3.3 70B Instruct by 110.911.9+1
- OpenBookQALlama 3.1 70B Instruct by 193.892.8+1
- GSM8Ktie93.894.2tie
- BFCL v2tie77.577.3tie
- MMLUtie8686tie
- MMLU (CoT) (Pass@1 with chain-of-thought)tie8686tie
Questions people ask
Which is better, Llama 3.1 70B Instruct or Llama 3.3 70B Instruct?
Llama 3.3 70B Instruct wins three of the six areas where both have results: reasoning, long documents and following instructions. Llama 3.1 70B Instruct wins facts, and costs 22% less. They are level on coding and agents.
Which is better for coding?
Neither. All 2 coding tests both models report are ties.
Which is cheaper?
Llama 3.1 70B Instruct costs $0.56 per million input tokens and $0.56 per million output tokens; Llama 3.3 70B Instruct costs $0.71 and $0.72. That makes Llama 3.1 70B Instruct about 22% cheaper for the same work.
How do you compare the two?
We use the 30 benchmark tests both models have published scores on. The verdict counts the 10 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 20 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.