Llama 3 Anyscale: pricing, context window and benchmarks
Llama 3 Anyscale is a multimodal model from Anyscale. It offers a 65,536-token context window, costs $3.94 per million input tokens and $19.76 per million output tokens, and averages 74.7 across the index's benchmark suite.
Pricing
Llama 3 Anyscale costs $3.94 per million input tokens and $19.76 per million output tokens. A blended 3:1 input-to-output mix works out to about $7.90 per million tokens. Cached input is listed at $0.49 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $3.94 | per 1M tokens |
| Output | $19.76 | per 1M tokens |
| Cached input | $0.49 | per 1M tokens |
| Blended 3:1 | $7.90 | estimate per 1M tokens |
| 1M in + 1M out | $23.70 | 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 | 77.5 | 257 | 57th | |
| GPQA Diamond | 70.6 | 245 | 59th | |
| SWE-bench Verified | 63.0 | 252 | 58th | |
| HumanEval | 79.8 | 238 | 60th | |
| MATH-500 | 78.8 | 241 | 60th | |
| AIME 2025 | 68.7 | 284 | 53th | |
| LiveCodeBench | 64.7 | 301 | 50th | |
| IFEval | 84.9 | 195 | 68th | |
| TruthfulQA | 77.3 | 272 | 55th | |
| HellaSwag | 85.9 | 218 | 64th | |
| MMMU | 70.8 | 278 | 54th | |
| Arena Elo | 1397.8 | 213 | 65th |
What Llama 3 Anyscale is good for
Llama 3 Anyscale 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 $3.94 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 $3.94 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
- No standout weakness in the indexed specifications, but real-world behaviour still needs testing.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Llama 3 Modal | Modal | 32768 | $3.82 | 74.9 |
| Llama 3 Anyscale | Anyscale | 65536 | $3.94 | 74.7 |
| Llama 3 Lambda Chat | Lambda | 32768 | $3.73 | 74.7 |
| Llama 3 Replicate Chat | Replicate | 200000 | $3.74 | 74.7 |
| Phi-4 Chat | Microsoft | 65536 | $3.66 | 74.7 |
| Llama 3 Anyscale Chat | Anyscale | 65536 | $3.67 | 74.6 |
FAQ
How much does Llama 3 Anyscale cost?
Llama 3 Anyscale is listed at $3.94 per million input tokens and $19.76 per million output tokens. Cached input is $0.49 per million tokens. A blended 3:1 input-to-output mix is about $7.90 per million tokens.
What is the context window of Llama 3 Anyscale?
Llama 3 Anyscale 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 Anyscale open weights?
Yes. Llama 3 Anyscale is listed as an open-weight model under a open weights licence.
How does Llama 3 Anyscale score on benchmarks?
Llama 3 Anyscale averages 74.7 across the scored benchmarks in this index, ranking #259 of 600 models. Its strongest relative result is Arena Elo at 1397.8.
What is Llama 3 Anyscale best for?
Llama 3 Anyscale 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 $3.94 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.