# Meta Llama 3.1 8B Instruct

As of October 22, 2026, Meta Llama 3.1 8B Instruct ranks #122 of 123 models overall with 17.9 ±7.1 on the LLMBenchmark Index. Llama 3.1 8B Instruct ranks #115 of 122 in reasoning, its strongest category. Meta Llama 3.1 8B Instruct costs $0.05 per million input tokens and $0.08 per million output tokens. Llama 3.1 8B Instruct was released on July 23, 2024.

Page: https://llmbenchmark.ai/models/llama-3-1-8b-instruct
Data from October 22, 2026. Every figure below names the board it was read from and the day.

- Maker: Meta Llama
- Released: July 23, 2024
- Price per million input tokens: $0.05
- Price per million output tokens: $0.08
- Context window: 131K tokens
- Overall rank: 122 of 123, by the LLMBenchmark Index
- LLMBenchmark Index: 17.9 ±7.1 (90% margin, from 4 of 7 inputs)
- Artificial Analysis Intelligence Index: 6.9

Benchmark results (51 of 195):

- AA Intelligence Index: 6.9 (source: Artificial Analysis (via OpenRouter), read October 22, 2026)
- AA Coding Index: 5.4 (source: Artificial Analysis (via OpenRouter), read October 22, 2026)
- AA Math Index: 4.3 (source: Artificial Analysis, read October 22, 2026)
- GPQA Diamond (Artificial Analysis run): 25.9% (source: Artificial Analysis, read October 22, 2026)
- Humanity's Last Exam (Artificial Analysis run): 5.3% (source: Artificial Analysis, read October 22, 2026)
- MMLU Pro (Artificial Analysis run): 47.6% (source: Artificial Analysis, read October 22, 2026)
- LiveCodeBench (Artificial Analysis run): 11.6% (source: Artificial Analysis, read October 22, 2026)
- Terminal-Bench Hard (Artificial Analysis run): 0.8% (source: Artificial Analysis, read October 22, 2026)
- IFBench: 28.6% (source: Artificial Analysis, read October 22, 2026)
- Long context reasoning (AA-LCR): 18.0% (source: Artificial Analysis, read October 22, 2026)
- τ²-Bench Telecom: 16.4% (source: Artificial Analysis, read October 22, 2026)
- AIME 2025 (Artificial Analysis run): 4.3% (source: Artificial Analysis, read October 22, 2026)
- WeirdML: 1.7% (source: Epoch AI benchmarking hub, read October 22, 2026)
- BALROG: 15.1% (source: Epoch AI benchmarking hub, read October 22, 2026)
- GPQA Diamond (Epoch AI run): 27.0% (source: Epoch AI benchmarking hub, read October 22, 2026)
- OTIS Mock AIME: 1.7% (source: Epoch AI benchmarking hub, read October 22, 2026)
- MATH level 5: 22.9% (source: Epoch AI benchmarking hub, read October 22, 2026)
- GSM8K: 82.4% (source: Epoch AI benchmarking hub, read October 22, 2026)
- MMLU: 56.1% (source: Epoch AI benchmarking hub, read October 22, 2026)
- DTBench: 50.9% (source: Epoch AI benchmarking hub, read October 22, 2026)
- LMCA: 5.4% (source: Epoch AI benchmarking hub, read October 22, 2026)
- PIQA: 81.2% (source: Epoch AI benchmarking hub, read October 22, 2026)
- Epoch Capabilities Index: 116.5 (source: Epoch AI benchmarking hub, read October 22, 2026)
- LMArena Text: 1187 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: coding: 1195 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: hard prompts: 1175 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: math: 1179 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: instruction following: 1159 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: creative writing: 1154 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: English: 1218 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: non-English: 1148 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Chinese: 1151 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: German: 1144 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Japanese: 1061 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Korean: 1053 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: French: 1178 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Spanish: 1169 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: Russian: 1158 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: multi-turn: 1172 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: longer queries: 1181 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: expert: 1144 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: hard prompts in English: 1196 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: software and IT: 1195 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: mathematical work: 1180 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: science: 1193 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: medicine and healthcare: 1182 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: legal and government: 1206 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: business and finance: 1166 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: writing and language: 1162 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- LMArena Text: entertainment and media: 1145 Elo (source: LMArena (Hugging Face dataset), read October 22, 2026)
- Berkeley Function Calling Leaderboard (BFCL): 25.8% (source: Berkeley Function Calling Leaderboard, read October 22, 2026)