Llama 3.1 Cerebras Base: pricing, context window and benchmarks
Llama 3.1 Cerebras Base is a open-weight model from Cerebras. It offers a 65,536-token context window, costs $3.36 per million input tokens and $15.59 per million output tokens, and averages 72.7 across the index's benchmark suite.
Pricing
Llama 3.1 Cerebras Base costs $3.36 per million input tokens and $15.59 per million output tokens. A blended 3:1 input-to-output mix works out to about $6.42 per million tokens. Cached input is listed at $0.32 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $3.36 | per 1M tokens |
| Output | $15.59 | per 1M tokens |
| Cached input | $0.32 | per 1M tokens |
| Blended 3:1 | $6.42 | estimate per 1M tokens |
| 1M in + 1M out | $18.95 | estimate |
Benchmarks
Scores below are modelled estimates on a 0–100 scale. Use them to compare relative position, not as official vendor results.
| Benchmark | Score | Rank | Percentile | Relative |
|---|---|---|---|---|
| MMLU-Pro | 77.4 | 265 | 56th | |
| GPQA Diamond | 66.9 | 332 | 45th | |
| SWE-bench Verified | 58.6 | 347 | 42th | |
| HumanEval | 79.2 | 259 | 57th | |
| MATH-500 | 74.2 | 342 | 43th | |
| AIME 2025 | 67.8 | 298 | 50th | |
| LiveCodeBench | 67.7 | 231 | 62th | |
| IFEval | 82.7 | 253 | 58th | |
| TruthfulQA | 75.1 | 329 | 45th | |
| HellaSwag | 82.6 | 332 | 45th | |
| MMMU | 67.9 | 342 | 43th | |
| Arena Elo | 1387.0 | 259 | 57th |
What Llama 3.1 Cerebras Base is good for
Llama 3.1 Cerebras Base is an open-weight model, so it can be self-hosted or run through a third-party inference provider. That matters for data residency, fine-tuning and long-run cost control. At $3.36 per million input tokens (hosted list price) and a 65,536-token context window, it is a practical option for teams that want more control than a closed API allows.
Strengths
- Input price of $3.36 per million tokens is at or below the index median.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
- Accepts more than text, including image input where the provider supports it.
Trade-offs
- Average benchmark score is below the index median; test it on your own task.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Mixtral Instruct | Mistral AI | 65536 | $3.35 | 72.9 |
| Grok 4 Nano | xAI | 16384 | $8.97 | 72.7 |
| Llama 3.1 Cerebras Base | Cerebras | 65536 | $3.36 | 72.7 |
| Mistral Small 3 Base | Mistral AI | 16384 | $3.19 | 72.7 |
| Yi-Lightning Base | 01.AI | 16384 | $3.31 | 72.7 |
| Mixtral Chat | Mistral AI | 65536 | $3.30 | 72.6 |
FAQ
How much does Llama 3.1 Cerebras Base cost?
Llama 3.1 Cerebras Base is listed at $3.36 per million input tokens and $15.59 per million output tokens. Cached input is $0.32 per million tokens. A blended 3:1 input-to-output mix is about $6.42 per million tokens.
What is the context window of Llama 3.1 Cerebras Base?
Llama 3.1 Cerebras Base has a context window of 65,536 tokens and can emit up to 16,384 output tokens per request in this index.
Is Llama 3.1 Cerebras Base open weights?
Yes. Llama 3.1 Cerebras Base is listed as an open-weight model under a open weights licence.
How does Llama 3.1 Cerebras Base score on benchmarks?
Llama 3.1 Cerebras Base averages 72.7 across the scored benchmarks in this index, ranking #309 of 600 models. Its strongest relative result is Arena Elo at 1387.0.
What is Llama 3.1 Cerebras Base best for?
Llama 3.1 Cerebras Base is an open-weight model, so it can be self-hosted or run through a third-party inference provider. That matters for data residency, fine-tuning and long-run cost control. At $3.36 per million input tokens (hosted list price) and a 65,536-token context window, it is a practical option for teams that want more control than a closed API allows.