Seed-OSS Instruct: pricing, context window and benchmarks
Seed-OSS Instruct is a multimodal model from ByteDance. It offers a 128,000-token context window, costs $3.79 per million input tokens and $20.28 per million output tokens, and averages 76.0 across the index's benchmark suite.
Pricing
Seed-OSS Instruct costs $3.79 per million input tokens and $20.28 per million output tokens. A blended 3:1 input-to-output mix works out to about $7.91 per million tokens. Cached input is listed at $0.36 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $3.79 | per 1M tokens |
| Output | $20.28 | per 1M tokens |
| Cached input | $0.36 | per 1M tokens |
| Blended 3:1 | $7.91 | estimate per 1M tokens |
| 1M in + 1M out | $24.07 | 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 | 78.5 | 226 | 62th | |
| GPQA Diamond | 72.3 | 215 | 64th | |
| SWE-bench Verified | 63.3 | 249 | 59th | |
| HumanEval | 80.2 | 228 | 62th | |
| MATH-500 | 81.2 | 187 | 69th | |
| AIME 2025 | 71.2 | 233 | 61th | |
| LiveCodeBench | 65.4 | 287 | 52th | |
| IFEval | 82.8 | 252 | 58th | |
| TruthfulQA | 79.3 | 207 | 66th | |
| HellaSwag | 87.2 | 169 | 72th | |
| MMMU | 74.5 | 170 | 72th | |
| Arena Elo | 1400.9 | 206 | 66th |
What Seed-OSS Instruct is good for
Seed-OSS Instruct 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 $3.79 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 $3.79 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 |
|---|---|---|---|---|
| Command R+ Chat | Cohere | 32768 | $4.02 | 76.0 |
| Gemini 3 Flash Mini | Google DeepMind | 200000 | $10.71 | 76.0 |
| Llama 3.1 Cerebras | Cerebras | 128000 | $4.01 | 76.0 |
| Llama 3.1 Cerebras Chat | Cerebras | 32768 | $4.07 | 76.0 |
| Seed-OSS Instruct | ByteDance | 128000 | $3.79 | 76.0 |
| Sonar | Perplexity | 200000 | $10.57 | 76.0 |
FAQ
How much does Seed-OSS Instruct cost?
Seed-OSS Instruct is listed at $3.79 per million input tokens and $20.28 per million output tokens. Cached input is $0.36 per million tokens. A blended 3:1 input-to-output mix is about $7.91 per million tokens.
What is the context window of Seed-OSS Instruct?
Seed-OSS Instruct has a context window of 128,000 tokens and can emit up to 16,384 output tokens per request in this index.
Is Seed-OSS Instruct open weights?
Yes. Seed-OSS Instruct is listed as an open-weight model under a open weights licence.
How does Seed-OSS Instruct score on benchmarks?
Seed-OSS Instruct averages 76.0 across the scored benchmarks in this index, ranking #224 of 600 models. Its strongest relative result is Arena Elo at 1400.9.
What is Seed-OSS Instruct best for?
Seed-OSS Instruct 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 $3.79 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.