GPT-4o mini vs Llama 3.1 Instruct 405B
Wins 0 of 4 areas
—
Wins 3 of 4 areas
Coding · Reasoning · Following instructions
Llama 3.1 Instruct 405B is the stronger all-rounder.
Scores updated · 26 tests both models report · How we compare
Where each one wins
Tests won in each of the four 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.
- CodingWriting and fixing software02Llama 3.1 Instruct 405B2 of 2 tests
- ReasoningHard problems that need careful thinking01Llama 3.1 Instruct 405B1 of 2 tests · 1 tie
- Following instructionsDoing exactly what it is asked01Llama 3.1 Instruct 405B1 of 1 test
- AgentsCarrying out multi-step tasks on its own00Even0 each · 1 tie
Agents 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.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Llama 3.1 Instruct 405B pulls ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+15.8points ahead
- Graduate-level biology, physics and chemistry questionsGPQA Diamond+8.9points ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+8points ahead
Where GPT-4o mini pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 26 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingLlama 3.1 Instruct 405B
- SWE-bench VerifiedLlama 3.1 Instruct 405B by 15.88.724.5+15.8
- SciCodeLlama 3.1 Instruct 405B by 722.929.9+7
ReasoningLlama 3.1 Instruct 405B
- GPQA DiamondLlama 3.1 Instruct 405B by 8.942.651.5+8.9
- Humanity's Last Examtie4.24tie
Following instructionsLlama 3.1 Instruct 405B
- IFBenchLlama 3.1 Instruct 405B by 83139+8
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.
- HarmBenchGPT-4o mini by 22.284.962.7+22.2
- MMLULlama 3.1 Instruct 405B by 21.866.888.6+21.8
- MBPPGPT-4o mini by 16.484.868.4+16.4
- GSM8KLlama 3.1 Instruct 405B by 12.584.396.8+12.5
- MMLU-ProLlama 3.1 Instruct 405B by 11.661.773.3+11.6
- LiveCodeBenchGPT-4o mini by 7.835.527.7+7.8
- SimpleQALlama 3.1 Instruct 405B by 7.29.917.1+7.2
- naturalquestions_closedbookLlama 3.1 Instruct 405B by 738.545.6+7
- MATHGPT-4o mini by 6.480.273.8+6.4
- DROPLlama 3.1 Instruct 405B by 5.179.784.8+5.1
- MGSMLlama 3.1 Instruct 405B by 4.68791.6+4.6
- Artificial Analysis Coding IndexLlama 3.1 Instruct 405B by 3.111.414.5+3.1
- AIR-Bench 2024Llama 3.1 Instruct 405B by 2.356.358.6+2.3
- OpenBookQALlama 3.1 Instruct 405B by 29294+2
- NarrativeQAGPT-4o mini by 1.976.874.9+1.9
- anthropic_red_teamGPT-4o mini by 1.898.396.5+1.8
- HumanEvalLlama 3.1 Instruct 405B by 1.887.289+1.8
- simple_safety_testsLlama 3.1 Instruct 405B by 197.898.8+1
- AA Intelligencetie6.77.3tie
- XSTesttie9695.9tie
Questions people ask
Which is better, GPT-4o mini or Llama 3.1 Instruct 405B?
Llama 3.1 Instruct 405B wins three of the four areas where both have results: coding, reasoning and following instructions. GPT-4o mini wins none. They are level on agents.
Which is better for coding?
Llama 3.1 Instruct 405B. It wins 2 of the 2 coding tests both models report; GPT-4o mini wins none.
How do you compare the two?
We use the 26 benchmark tests both models have published scores on. The verdict counts the 6 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.