GPT-NeoX: pricing, context window and benchmarks
GPT-NeoX is a open-weight model from EleutherAI. It offers a 16,384-token context window, costs $1.57 per million input tokens and $7.57 per million output tokens, and averages 61.7 across the index's benchmark suite.
Pricing
GPT-NeoX costs $1.57 per million input tokens and $7.57 per million output tokens. A blended 3:1 input-to-output mix works out to about $3.07 per million tokens. Cached input is listed at $0.15 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $1.57 | per 1M tokens |
| Output | $7.57 | per 1M tokens |
| Cached input | $0.15 | per 1M tokens |
| Blended 3:1 | $3.07 | estimate per 1M tokens |
| 1M in + 1M out | $9.14 | 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 | 67.6 | 422 | 30th | |
| GPQA Diamond | 54.6 | 468 | 22th | |
| SWE-bench Verified | 46.4 | 449 | 25th | |
| HumanEval | 64.9 | 442 | 26th | |
| MATH-500 | 64.0 | 425 | 29th | |
| AIME 2025 | 55.8 | 411 | 32th | |
| LiveCodeBench | 52.7 | 418 | 30th | |
| IFEval | 69.8 | 464 | 23th | |
| TruthfulQA | 67.6 | 422 | 30th | |
| HellaSwag | 75.0 | 442 | 26th | |
| MMMU | 60.5 | 424 | 29th | |
| Arena Elo | 1277.8 | 448 | 25th |
What GPT-NeoX is good for
GPT-NeoX 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 $1.57 per million input tokens (hosted list price) and a 16,384-token context window, it is a practical option for teams that want more control than a closed API allows.
Strengths
- Input price of $1.57 per million tokens is at or below the index median.
- Open weights make self-hosting, fine-tuning and data-residency review possible.
Trade-offs
- Context window of 16,384 tokens is small for long-document work.
- Average benchmark score is below the index median; test it on your own task.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Sonar Nano | Perplexity | 4096 | $4.37 | 61.9 |
| GLM-4.6 7B | Zhipu AI | 16384 | $1.52 | 61.8 |
| GPT-NeoX | EleutherAI | 16384 | $1.57 | 61.7 |
| MiniMax-M2 8B | MiniMax | 32768 | $1.65 | 61.7 |
| Auto Mini | OpenRouter | 4096 | $4.24 | 61.6 |
| Mistral Large 3 8B | Mistral AI | 16384 | $1.62 | 61.6 |
FAQ
How much does GPT-NeoX cost?
GPT-NeoX is listed at $1.57 per million input tokens and $7.57 per million output tokens. Cached input is $0.15 per million tokens. A blended 3:1 input-to-output mix is about $3.07 per million tokens.
What is the context window of GPT-NeoX?
GPT-NeoX has a context window of 16,384 tokens and can emit up to 4,096 output tokens per request in this index.
Is GPT-NeoX open weights?
Yes. GPT-NeoX is listed as an open-weight model under a open weights licence.
How does GPT-NeoX score on benchmarks?
GPT-NeoX averages 61.7 across the scored benchmarks in this index, ranking #434 of 600 models. Its strongest relative result is Arena Elo at 1277.8.
What is GPT-NeoX best for?
GPT-NeoX 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 $1.57 per million input tokens (hosted list price) and a 16,384-token context window, it is a practical option for teams that want more control than a closed API allows.