Meta: Llama 3.1 8B Instruct
Provider: meta-llama · meta-llama/llama-3.1-8b-instruct
How Meta: Llama 3.1 8B Instruct pricing compares
At $0.08 / 1M output tokens, Meta: Llama 3.1 8B Instruct is the 1st-cheapest of 8 meta-llama models we track. It's cheaper than 98% of the 392 paid models on computetrail (7th-cheapest overall). Output costs 1.6× its input price ($0.05 / 1M in). For a typical 3:1 input-to-output workload that works out to about $0.06 / 1M blended. Reading cached input costs 50% less than fresh input, so prompt caching cuts repeat-context cost sharply. Its 131,072 tokens context window ranks 319th-largest of 392 models with a published limit.
Output price history
Across 60 daily snapshots since 2026-06-29, the output price has risen 166.7% — from $0.03 to $0.08 / 1M.
Similar-priced models
| Model | Provider | Output $/1M | Context |
|---|---|---|---|
| Mistral: Ministral 8B | mistralai | $0.11 | 128,000 |
| Mistral: Mistral Nemo | mistralai | $0.03 | 131,072 |
| Microsoft: Phi 4 | microsoft | $0.14 | 16,384 |
| Amazon: Nova Micro 1.0 | amazon | $0.14 | 128,000 |
| Qwen: Qwen3.5-9B | qwen | $0.15 | 262,144 |
| Cohere: Command R7B (12-2024) | cohere | $0.15 | 128,000 |
Prices are recorded daily from public provider data via OpenRouter and normalized to USD per 1M tokens. See methodology.