Llama 3 DeepInfra: pricing, context window and benchmarks
Llama 3 DeepInfra is a multimodal model from DeepInfra. It offers a 128,000-token context window, costs $3.67 per million input tokens and $14.87 per million output tokens, and averages 74.0 across the index's benchmark suite.
Pricing
Llama 3 DeepInfra costs $3.67 per million input tokens and $14.87 per million output tokens. A blended 3:1 input-to-output mix works out to about $6.47 per million tokens. Cached input is listed at $0.33 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $3.67 | per 1M tokens |
| Output | $14.87 | per 1M tokens |
| Cached input | $0.33 | per 1M tokens |
| Blended 3:1 | $6.47 | estimate per 1M tokens |
| 1M in + 1M out | $18.54 | 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 | 259 | 57th | |
| GPQA Diamond | 69.2 | 290 | 52th | |
| SWE-bench Verified | 60.2 | 323 | 46th | |
| HumanEval | 79.1 | 263 | 56th | |
| MATH-500 | 75.8 | 304 | 49th | |
| AIME 2025 | 67.7 | 299 | 50th | |
| LiveCodeBench | 64.9 | 295 | 51th | |
| IFEval | 81.4 | 290 | 52th | |
| TruthfulQA | 78.2 | 250 | 58th | |
| HellaSwag | 85.7 | 225 | 63th | |
| MMMU | 74.1 | 184 | 69th | |
| Arena Elo | 1370.4 | 314 | 48th |
What Llama 3 DeepInfra is good for
Llama 3 DeepInfra 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 $3.67 per million input tokens (hosted list price) and a 128,000-token context window, it is a practical option for teams that want more control than a closed API allows.
Strengths
- Large 128,000-token context window for long documents and agent traces.
- Input price of $3.67 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 |
|---|---|---|---|---|
| GPT-5.1 Nano | OpenAI | 32768 | $9.56 | 74.1 |
| Gemma 3 Instruct | Google DeepMind | 128000 | $3.72 | 74.1 |
| Reka Flash | Reka AI | 128000 | $9.95 | 74.1 |
| Hunyuan-Large Base | Tencent | 200000 | $3.64 | 74.0 |
| Llama 3 DeepInfra | DeepInfra | 128000 | $3.67 | 74.0 |
| Llama 3.3 Base | Meta AI | 32768 | $3.63 | 74.0 |
FAQ
How much does Llama 3 DeepInfra cost?
Llama 3 DeepInfra is listed at $3.67 per million input tokens and $14.87 per million output tokens. Cached input is $0.33 per million tokens. A blended 3:1 input-to-output mix is about $6.47 per million tokens.
What is the context window of Llama 3 DeepInfra?
Llama 3 DeepInfra has a context window of 128,000 tokens and can emit up to 8,192 output tokens per request in this index.
Is Llama 3 DeepInfra open weights?
Yes. Llama 3 DeepInfra is listed as an open-weight model under a open weights licence.
How does Llama 3 DeepInfra score on benchmarks?
Llama 3 DeepInfra averages 74.0 across the scored benchmarks in this index, ranking #279 of 600 models. Its strongest relative result is Arena Elo at 1370.4.
What is Llama 3 DeepInfra best for?
Llama 3 DeepInfra 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 $3.67 per million input tokens (hosted list price) and a 128,000-token context window, it is a practical option for teams that want more control than a closed API allows.