DeepSeek-R1 8B: pricing, context window and benchmarks
DeepSeek-R1 8B is a small open-weight model from DeepSeek. It offers a 16,384-token context window, costs $1.87 per million input tokens and $8.53 per million output tokens, and averages 64.3 across the index's benchmark suite.
Pricing
DeepSeek-R1 8B costs $1.87 per million input tokens and $8.53 per million output tokens. A blended 3:1 input-to-output mix works out to about $3.54 per million tokens. Cached input is listed at $0.20 per million tokens.
| Item | Price | Notes |
|---|---|---|
| Input | $1.87 | per 1M tokens |
| Output | $8.53 | per 1M tokens |
| Cached input | $0.20 | per 1M tokens |
| Blended 3:1 | $3.54 | estimate per 1M tokens |
| 1M in + 1M out | $10.40 | 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 | 67.2 | 429 | 29th | |
| GPQA Diamond | 58.0 | 426 | 29th | |
| SWE-bench Verified | 53.3 | 394 | 34th | |
| HumanEval | 69.8 | 397 | 34th | |
| MATH-500 | 67.0 | 407 | 32th | |
| AIME 2025 | 58.4 | 397 | 34th | |
| LiveCodeBench | 52.4 | 421 | 30th | |
| IFEval | 73.1 | 414 | 31th | |
| TruthfulQA | 67.6 | 419 | 30th | |
| HellaSwag | 77.9 | 403 | 33th | |
| MMMU | 62.7 | 402 | 33th | |
| Arena Elo | 1313.8 | 401 | 33th |
What DeepSeek-R1 8B is good for
DeepSeek-R1 8B is a small model. That makes it a good fit for classification, extraction, routing, summarization and other high-volume jobs where latency and cost dominate. It is not the right pick for open-ended frontier reasoning, but at $1.87 per million input tokens it can handle a large share of production traffic.
Strengths
- Input price of $1.87 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
- 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.
- Less reliable on open-ended reasoning and complex multi-step instructions.
How it compares
| Model | Provider | Context | Input / 1M | Avg score |
|---|---|---|---|---|
| Kimi K2 8B | Moonshot AI | 8192 | $1.86 | 64.7 |
| StableLM 2 Chat | Stability AI | 8192 | $1.79 | 64.6 |
| DeepSeek-R1 8B | DeepSeek | 16384 | $1.87 | 64.3 |
| Embed 4 Turbo | Cohere | 4096 | $5.13 | 64.0 |
| Embed 4 | Cohere | 32768 | $4.79 | 63.8 |
| SmolLM3 Instruct | Hugging Face | 8192 | $1.69 | 63.6 |
FAQ
How much does DeepSeek-R1 8B cost?
DeepSeek-R1 8B is listed at $1.87 per million input tokens and $8.53 per million output tokens. Cached input is $0.20 per million tokens. A blended 3:1 input-to-output mix is about $3.54 per million tokens.
What is the context window of DeepSeek-R1 8B?
DeepSeek-R1 8B has a context window of 16,384 tokens and can emit up to 16,384 output tokens per request in this index.
Is DeepSeek-R1 8B open weights?
Yes. DeepSeek-R1 8B is listed as an open-weight model under a open weights licence.
How does DeepSeek-R1 8B score on benchmarks?
DeepSeek-R1 8B averages 64.3 across the scored benchmarks in this index, ranking #409 of 600 models. Its strongest relative result is Arena Elo at 1313.8.
What is DeepSeek-R1 8B best for?
DeepSeek-R1 8B is a small model. That makes it a good fit for classification, extraction, routing, summarization and other high-volume jobs where latency and cost dominate. It is not the right pick for open-ended frontier reasoning, but at $1.87 per million input tokens it can handle a large share of production traffic.